- Email: [email protected]
Prototropic Tautomerism of Heteroaromatic Compounds: III. Five-Membered Rings and One Hetero Atom*
Prototropic Tautomerism of Heteroaromatic Compounds : I11. Five-Membered Rings and One Hetero AtomX
. .
A R KATRITZKY? University Chemical Laboratory. Cambridge. England AND
. .
J M LAGOWSKI Genetics Foundation. The U i h e r s i f y of Texas. Artstin. Texas
I . Tautomerism of Pyrroles Not Involving the Functional Group I1. Compounds with a Potential Hydroxyl Group . . . . . A . Hydroxyfurans . . . . . . . . . . . . . . B. Potential Dihydroxyfurans . . . . . . . . . . . C . a-Hydroxythiophenes . . . . . . . . . . . . D . P-Hydroxythiophenes . . . . . . . . . . . . E . a-Hydroxypyrroles . . . . . . . . . . . . . F. p-Hydroxypyrroles . . . . . . . . . . . . . . G. Poly-0x0- and -hydroxy-pyrroles and -indoles . . . . . H . Monohydroxyindoles . . . . . . . . . . . . I . Other Compounds with Potential Hydroxyl Groups . . . I11. Compounds with Potential Mercapto Groups . . . . . . IV . Compounds with Potential Amino Groups . . . . . . . A . Aminofurans . . . . . . . . . . . . . . . B . Aminothiophenes . . . . . . . . . . . . . . C . Aminopyrroles . . . . . . . . . . . . . . . D . Aminoindoles . . . . . . . . . . . . . . . E . Potential Amino Derivatives of Isoindole . . . . . . V . Compounds with Potential Methyl Groups . . . . . . V I . Other Substituted Pyrroles . . . . . . . . . . . . A . Vinylpyrroles . . . . . . . . . . . . . . . B . Nitrosopyrroles . . . . . . . . . . . . . . Errata . . . . . . . . . . . . . . . . . .
.
3 . 5 . 5 . 6 . 8 . 9 . 1 1 . 1 4 . 15 . 1 8 . 19 . 2 0 . 20 . 2 1 . 2 2 . 2 2 . 2 3 . 2 4 . 24 . 25 . 25 . 2 6 . 26
The most important potentially tautomeric thiophenes and furans are those carrying hydroxyl. mercapto. and amino groups. I n these compounds a prototropic shift can occur between the functional group *The first two chapters in this volume conclude the series of four articles on prototropic tautomerism; the first two articles appeared in Volume 1. Cross references to these articles include. for easy identification. the roman numeral given in the title . t Present address : School of Chemistry. University of East Anglia. Norwich. England . 1
2
A. R. KATRITZKY AND J. M. LAGOWSKI
and a ring carbon atom, cf. type ii described in Volume 1, Section I,A, of article I by Katritzky and Lagowski. A single functional group in the a-position gives rise to three possible tautomeric structures, 1-3, whereas two structures, 4 and 5, are possible with one functional group in the P-position. Similar structures are possible for the analogous pyrroles, but pyrroles which do not carry a substituent on the hetero nitrogen atom can undergo additional types of tautomerism (see later). The proportion of the various forms present a t equilibrium is dependent upon their relative stability. Forms 1 and 4 are “aromatic,” and, since the aromatic stabilization decreases in the order thiophene > pyrrole > furan, forms of this type would be expected to be the most favored for thiophenes. The nonaromatic forms are, however, also mesomeric as indicated by curly arrows in 2, 3, and 5.
“1
Pl
[31
Systems 2 and 3 are cross-conjugated, but 5 is not, and i t might have been expected that a-substituted compounds would be more prone to exist in the aromatic form 1 than the ,&compounds to exist as 4. From the limited evidence available, the reverse appears to be the case. When the hetero atom is not a very strong electron donor, i.e., 2 = S or 0, structure 3 would be expected to be relatively more stable than 2, and this is supported by the evidence available. When 2 = NR, the difference in the stabilities of 2 and 3 could be smaller. The nature of the substituent group X plays an important role in determining the relative stability of the various tautomeric forms. I n aliphatic systems the tendency of :C=C-XH to become :CH-C=X increases markedly in the order X = CH, < NH < 0, and certainly hydroxy compounds show less inclination to exist as such than do amino compounds. The position of S in this series is not completely
TAUTOMERISM: 111. &MEMBERED RINGS-1
HETERO ATOM
3
clear (see discussion in Section I11 of article 11, Volume 1 ) ; however, the thione-thiol equilibrium 6 + 7 appears to be more in favor of 7 than in the case of the corresponding keto-enol equilibrium.’ In agreement with this contention, mercaptothiophenes are more stable than the hydroxythiophenes (discussed later). Electron-withdrawing substituents stabilize the aromatic forms 1 and 4, as would be expected, by acting as an “electron sink” and by forming intramolecular hydrogen bonds with the XH group when this is sterically possible. On the other hand, a benzene ring fused in the 3,4-position greatly stabilizes 8 as compared to 9. The effect of a benzene ring fused in the 4,5-position is smaller, but probably also causes some preferential stabilization of the nonaromatic forms corresponding to 2 and 5. R-C-CCH,CO,Et
II S
=R-CCCH-C0,Et I SH
I. Tautomerism of Pyrroles Not Involving the Functional Group The tautomeric forms 11 and 12, called pyrrolenines, have often been postulated for pyrrole (lo), but there is no conclusive evidence for their existence.2 The report3 that two isomers exist as the pyrrolenine forms 13 and 14 (Ar = 3,4-dimethoxyphenyl) must be regarded with considerable doubt. Chemical evidence led to the conclusion that the conjugate acids of pyrrole probably exist predominantly as 16 or 17 rather than as 15.2,4,5Although infrared spectra were initially interpreted on the ‘ Z . Reyes and R. M. Silverstein, J . Am. Chem. SOC.80, 6367 (1958). ‘ A . Treibs and G. Fritz, Ann. Chem. Liebigs 611, 162 (1958). *D. A. Guthrie, A. W. Frank, and C. B. Purves, Can. J . Chem. 33, 729 (1955). ‘A.Treibs and K . H. Michl, Ann. Chem. Liebigs 577, 129 (1952). ‘A. Treibs and A. Ohorodnik, Ann. Chem. Liebigs 611, 139 (1958).
4
A. R. KATRITZKY AND J . M. LAGOWSKI
H
1131
~ 4 1
basis of structure 15: nuclear magnetic resonance spectra later proved that structures of type 16 were favored in ~ o l u t i o n Similarly, .~ ultraviolets and nuclear magnetic resonance spectras and basicity measurem e n t ~show ~ ~ that indole cations are formed by protonation a t the
p-position to give 18, substantiating the postulation which had been offered earlier on theoretical g r o ~ n d s . ~ Dipyrromethenes rapidly exchange the hydrogen atom attached to nitrogen, and the two isomers of unsymmetrical compounds (e.g., 19 and 20) cannot be separately isolated.lDDipyrromethenes form meso-
'E. Bullock, Can. J . Chem. 36, 1686 (1958). 'R. J. Abraham, E. Bullock, and S. S. Mitra, Can. J . Chem. 37, 1859 (1959). * M. J. Kamlet and J. C. Dacons, J . Org. Chem. 26, 220 (1961). OR.L. Hinnar and J. Lang, Tetrahedron, Letters No. 21, 12 (1960).
"G.Berti, A. Da Sett,imo, and D. Segnini, Gnzz. chim. ital.
91, 571 (1961).
1°H. Fischer and B. Walach, Ann. Chem. Liebigs 450, 109 (1926).
TAUTOMERISM : 111. &MEMBERED RINGS-1
n
HETERO ATOM
5
n
I
P I
meric cations (cf. 21) by proton addition a t n i t r ~ g e n whereas ,~ olefinic derivatives such as 22 undergo proton addition on carbon. 11. Compounds with a Potential Hydroxyl Group Heterocyclic compounds carrying hydroxyl groups may be compared with phenols. Thomsonll has reviewed the tautomeric behavior of phenols; often both tautomeric forms of polycyclic compounds such as naphthols can be isolated. Early work on hydroxythiophenes and -furans was also reviewed by Thomson,'l but until recently their chemistry has been in a somewhat confused state. A pattern is now beginning to emerge, a t least for the a-substituted con~pounds,which appear to exist as a3-ox0 derivatives and to attain equilibrium slowly with the corresponding a4-0x0 forms. For the a-hydroxy compounds, the equilibrium generally favors the a3-ox0 form.
A. HYDROXYFURANS Unsaturated 7-lactones, e.g., a- (23) and ,&angelica lactone (24), are well known. Compounds 23 and 24 are both converted by alkaline catalysts into an equilibrium mixture in which 23 predominates, the amount of the hydroxy form (25) present a t equilibrium being exceedingly small. True a-hydroxyfurans are unknown, and, although the preparation of both a- and ,f3-hydroxyfurans has been rep0rted,1~J~ these claims have often been refuted (see, e.g., reference 14). Me a 0
Me
Me
EL
"R. H. Thomson, Quart. Revs. (London) LO, 27 (1956). "H. H. Hodgson and R. R. Davies, J. Chem. SOC.p. 806 (1939). "H. H. Hodgson and R. R. Davies, J . Chem. SOC.p. 1013 (1939). I'M. P. Cava, C. L. Wilson, and C. J. Williams, J . A m . Chem. Soc. 78, 2303 (1956).
6
A. R. KATRITZKY AND
J. M. LAGOWSKI
Very little is known concerning the simple, monocyclic 3-hydroxyfurans (cf. reference 15). Both the 0x0 and hydroxy forms of the substituted 3-hydroxyfurans 26 and 27 (R = H, C,H,) have been isolated>G,li but the individual tautomers slowly undergo interconversion. The enol forms give a positive reaction with ferric chloride, react rapidly with bromine, and form a peroxide with oxygen. From chemical evidence, the benzo derivatives of 3-hydroxyfuran, 28lS and 29,19appear to exist predominantly in the 0x0 form, and this is further supported by ultraviolet spectral data.20 Stefanye and Howard?’
P I
~ 9 1
concluded from infrared spectroscopic data that 5,7-dichloro-3-hydroxybenzofuran exists in the 0x0 form, but that the hydroxy form of its 2- (5’,7’-dichlorobenzofuryl) derivative, i.e., 3-hydroxy-5,5’,7,7’tetrachloro-2,3’-bibenzofuran, apparently predominates.
B.
POTENTIhL
DIHTDROXYFURANS
a-Keto-y-lactones appear to exist in the dioxo form 30,22 but enolization can occur when this leads to extended conjugation. The I5E. VotoEek and S. Malachta, Collection Czechoslov. Chem. Communs. 4, 87 ( 1932). leE. P. Kohler, F. H. Westheimer, and M. Tishler, J . A m . Chem. Sac. 58, 264 (1936). “ E . P.Kohler and D. W. Woodward, J. A m . Chem. SOC.58, 1933 (1936). “ K . v. Auwers and E. Auffenberg, Ber. deut. chem. Ges. 52, 92 (1919). ”P. Emmott and R. Livingstone, J. Chem. SOC.p. 4629 (1958). *Mme. Ramart-Lucas and M. van Cowenbergh, Bull. sac. chim. France p. 1381 (1935). ‘ID. Stefanye and W. L. Howard, J . 01.8.Chem. 20, 813 (1955). P1. A. Plattner and L. M. Jampolsky, Helv. Chim. Acta 26, 687 (1943).
TAUTOMERISM : 111. 5-MEMBERED mNGs-1
HETERO ATOM
7
enol forms of 31 (R = C,H,) and 31 (R = H) have been shown to be predominant by chemical evidencez3 and by infrared spectral datalZ4
~301
[311
respectively. Tetronic acids exist predominantly in the dioxo form (32) in solvents of low polarity, while the existence of the monoenol form (33) has been established in other solvents by infraredz5 and ultraviolet spectral comparisonsz6 and from dipole moment dat.a.27Haynes and Plimmerz7ahave recently reviewed the structure of these compounds [see also reference 28(a)], and the tautomerism of vitamin A (U), which has a related structure, has also been surveyed.z8(b)Analogous compounds carrying an amino group in the 3-position are also known.28(c)
Succinic anhydride (35, Z = 0) can theoretically tautomerize to 36, but all the evidence indicates that i t exists overwhelmingly as 35; for example, the infrared spectrum shows v C=O 23W. E. Bachmann, G. I. Fujimoto, and L. B. Wick, J. Am. Chem. SOC.72, 1995 (1950).
*’L. Mangoni and M . Belardini, Ann. chim. (Rome) 50, 322 (1960). “ L . A. Duncanson, J. Chem. SOC.p. 1207 (1953). “ E . R. H. Jones and M. C. Whiting, J. Chem. Soc. p. 1419 (1949). ” W . D. Kumler, J. Am. Chem. Soc. 62, 3292 (1940). *“L. J. Haynes and J. R. Plimmer, Quart. Revs. (London) 14, 292 (1960). “H. von Euler and B. Eistert, “Chemie und Biochemie der Reduktone und Reduktonate,” (a) p. 159, (b) p. 185, (c) p. 261. F. Enke, Stuttgart, Germany, 1957.
H. M. Randall, R. G. Fowler, N. Fuson, and J. R. Dangl, “Infrared Determination of Organic Structures.’’ D. Van Nostrand, New York, 1949.
8
A. R. RATRITZKY AND J . M. LAGOWSKI
Infrared and nuclear magnetic resonance spectral evidence led Kendall and Hajos3" to conclude that furan-3,4-dione (37) exists as such, which is surprising since cyclic a-diketones with five-membered rings are usually monoenolized. The 2,bdicarbethoxy derivative 38 was earlier stated to exist in the dihydroxy form.31 0a 0H
O
OOH
,OH
H\ O
V
[371
0
CO,E t
Et0,C [381
C. a-HYDROXYTHIOPHENES The infrared spectrum of 2-hydroxythiophene was originally interpreted as showing both v O H and vC=O peaks indicating that it exists as a mixture of the hydroxy form 39 (R = H ) and a t least one of the 0x0 forms, 40 and/or 41.32The ultraviolet spectrum of 2-hydroxythiophene is different from that of the corresponding methyl ether (39, R = Me) suggesting the presence of the chromophore contained in structure 40. The facts that this compound gives a positive
color test with ferric chloride, is a weak acid, and undergoes reactions characteristic of a phenolic hydroxyl group have been advanced as further evidence for the presence of the hydroxy form.32The infrared % E .C . Kendall and 2. G. Hajos, J. Am. Chem. Soc. 82, 3219 (1960). 3i W.H. Hoehn, Iowa State Coll. .I. Sci. 11, 66 (1936); Chem. Abstr. 31, 1800 (1937). "C.D.Hurd and K. L. Kreuz, J. Am. Chem. SOC.72, 5543 (1950).
TAUTOMERISM : 111. &MEMBERED RINGS-1
HETERO ATOM
9
spectrum of 5-phenyl-2-hydroxythiophene measured in chloroform solution shows a C=O absorption band, but its ultraviolet spectrum in ethanol is rather similar to that of the corresponding 2-methoxy derivative. Comparison of the ultraviolet spectra was complicated by a spontaneous oxidation of the latter compound. The ultraviolet spectrum measured in chloroform or isooctane did not correspond to that obtained in ethanol. Chemical evidence was also considered to support a mobile keto-enol tautomeric e q ~ i l i b r i u m . ~ ~ The foregoing conclusions must now be modified on the basis of a recent, detailed investigation of the tautomerism of 2-hydroxy5-methylthiophene by Gronowite and Hoffman3* using nuclear magnetic resonance and infrared spectroscopy. Compounds 42 and 43 form an equilibrium mixture containing 85% of 43. However, equilibrium is attained slowly, and 42 can be obtained essentially pure by dissolving the mixture in a base and precipitating with acid, whereas almost pure 43 can be isolated by distillation. The effect of the methyl group in stabilizing structure 42 is illustrated by the fact that 2hydroxythiophene itself exists essentially completely as 40.34
and it 2-Hydroxythianaphthene has been isolated in two has been suggested that these may be the individual keto and enol tautomers.ll A reinvestigation of this system using modern techniques would be welcome.
D. /3-HYDROXYTHIOPHENES The infrared spectrum of 3-hydroxythiophene has been interpreted by Ford and M a ~ k a to y ~show ~ that it exists as a mixture of both the hydroxy (44) and the 0x0 forms (45). 5-Phenyl-3-hydroxythiophene apparently behaves similarly to the 2-hydroxy isomer, the ultraviolet asA. I. Kosak, R. J. F. Palchak, W. A. Steele, and C. M. Selwitz, J . Am. Chem. SOC.76, 4450 (1954). Gronowitz and R. A . Hoffman, Arkiv Kemi 15, 499 (1960). " C. Marschalk, J . prakt. Chem. 88, 227 (1913). "M. C. Ford and D. Mackay, J . Chem. SOC.p. 4985 (1956).
10
A. R. KATRITZKY AND J. M. LAGOWSKI
spectrum measured in ethanol suggesting that it exists in the hydroxy form, whereas that determined in chloroform indicates that one of the 0x0 forms predon~inates.~~ An alkoxycarbonyl group in the 2position of the thiophene nucleus probably stabilizes a hydroxyl group in the 3-position (cf. reference 37). I n 1920, Auwers and T h i e ~ suggested ,~ that 3-hydroxythianaphthene
might exist as 46 in the solid &ate, but partly as 47 in alcoholic solution. These conclusions appear to be supported in part by more recent although infrared spectral data indicate that the 0x0
form (46) greatly predominates in both chloroform and carbon disulfide solutions.4o Snyder and his c o - ~ o r k e r s ~ ' assigned ~*~ structures 48 and 49 to these p-hydroxythiophene derivatives on the basis of chemical evidence and infrared and nuclear magnetic resonance spectral data. Infrared and nuclear magnetic resonance spectra further indicate that compounds of type 49 exist as dimers, probably hydrogen bonded, when R = OC,H, or CH,, but as monomeric enols when R = H.', arH.Fiesselmann and P. Schipprak, Chem. Ber. 89, 1897 (1956). 'K. v. Auwers and W. Thies, Ber. deut. chem. Ges. 53, 2285 (1920). *'F. Krohnke, Chem. Ber. 92, cxiv (1959). J. Holt, A. E. Kellie, D. G. O'Sullivan, and P. W. Sadler, J. Chem. SOC. p. 1217 (1958).
W. Carpenter and H. R. Snyder, J. Am. Chem. SOC.82, 2592 (1960). '*D. S. Matteson and H. R. Snyder, J . Org. Chem. 22, 1500 (1957). r 3 R .J. Tuite, A. D. Josey, and H. R. Snyder, J. Am. Chem. Soc. 82, 4360
"
( 1960).
TAUTOMERISM : 111. &MEMBERED
RINGS-1
HETERO ATOM
11
On the basis of chemical evidence, Swiss investigators have postulated that 3,4-dihydroxythiophenes exist as diols, i.e., as 50.44145 H
H
E. a-HYDROXYPYRROLES 2-Hydroxypyrroles are thought to exist in 0x0 forms such as 51 or 52; structure 53 illustrates a third possible 0x0 form.46 Chemical evidence for tautomerism in the hydroxypyrroles has been reviewed by Fischer and Orth.47 Since 2-hydroxypyrrole itself is unstable and Q-0I
H [511
0
I H [521
0
M?AyA HCI CO,R
0
[531
rapidly resinifies, studies have been confined to some of its more stable derivatives. The ultraviolet spectrum of 54 (R = C0,Et) is different from that of 55, and 54 (R = C0,Et) does not give a positive test with ferric ~ h l o r i d e , ~which ~ , ~ ~led to its formulation as shown. On the basis of their infrared (and ultraviolet) spectra, compounds 56 (R = H),5O 56 (R = A c ) , ~ "and 5751 must exist in 0x0 forms since they exhibit v C=O absorption bands: 56 (R = H) a t 5 . 9 5 ~(1681 cm-I), 57 a t 5 . 9 0 ~(1695 cm-l), and 56 (R = Ac) two bands a t 5.87 and 5 . 9 5 ~(1704 and 1681 cm-'). Proton resonance " P . Karrer and F. Kehrer, Helv. Chim. Acta 27, 142 (1944). "P. Karrer, R. Keller, and E. Usteri, Helv. Chim. Acta 27, 237 (1944). "H. Lapin and A. Horeau, Bull. SOC. chim. France p. 1703 (1960). " H.Fischer and H. Orth, "Die Chemie des Pyrrols." Vol. I, p. 124.Akademische Verlag., Leipzig, 1934. "C. A. Grob and P. Ankli, Helv. Chim. Acta 32, 2010 (1949). "C.A. Grob and P. Ankli, Helv. Chim. Acta 32, 2023 (1949). 'OH.Plieninger and M . Decker, Ann. Chem. Liebigs 598, 198 (1956). 61K. E. Schulte, J. Reisch, and R. Hobl, Arch. Pharm. 293, 687 (1960).
12
A. R. KATRITZKY AND J. M. LAGOWSKI
bh
R
spectra confirm structure 56 (R = H).51aThe A4-structure was assigned to 57 because it added water reversibly, and this is considered to be a characteristic reaction of A4-pyrrolones,52whereas the A3structure was assigned to 56 (R = H) because the ultraviolet spectrum appeared to be of the crotonic acid type.5o Ultraviolet spectra suggested, and nuclear magnetic resonance spectra proved, that 58 (R = H, OH) should be assigned the A3-structure The tautoineric behavior of compounds of type 59 has been discussed by Meyer and V a ~ g h a n .An ~ ~ intramolecularly hydrogenbonded 0x0 structure has been assigned to 60 on the basis of its infrared spectrum;5 whereas unambiguous chemical evidence, i.e., ozonolysis to succimide, confirmed the isolation of 61 in the 0x0 form.56 The foregoing results may be summarized as follows: potential a-hydroxypyrroles exist as pyrrolones. Substituents in the 3-position and in the 5-position favor the A3- and the A4-pyrrolone structure, respectively, as is to be expected. For 3,4,5-trisubstituted compounds, such as 61b, the A3-structure appears to be preferred.51aAn electron&'OH. Plieninger, H. Bauer, and A. R. Katritzky, Ann. Chem. Liebigs 654, 165 ( 1962).
"K,E. Schulte and J. Reisch, Arch. Phnrm. 292, 51 (1959).
53J. A. Moore and J. Binkert, J . A m . Chem. SOC.81, 6029 (1959). "W. L. Meyer and W. R . Vaughan, J. Org. Chem. 22, 1565 (1957). "P. L. Southwick and R. J. Owellen, J. Org. Chem. 25, 1133 (1960). uJ. A. Elvidge, J. S. Fitt, and R. P. Linstead, J. Chem. SOC.p. 235 (1956).
TAUTOMERISM : 111. &MEMBERED RINGS-1
HETERO ATOM
13
0
H
-NR
accepting substituent in the 4-position favors the A4-pyrrolone structure, because conjugation between the substituent and the cyclic nitrogen atom is then possible. An electron-accepting substituent in the 5-position might be expected to stabilize the hydroxypyrrole form, because only then is extended conjugation possible in the molecule: the aldehyde 61a does, indeed, exist as shown on proton resonance spectral
Me)rr7Me
OHC
N
Me\
,Me
OH
Interesting tautomeric possibilities exist in the xanthobilirubic acid series (cf. reference 57) which can be illustrated by the equilibrium 62$63. More complex examples of the same type are found among the linear tetrapyrrole pigments-the bilenes, bilidienes, and bilitrienes-and have been discussed by Stevens.58 Relatively little evidence is available concerning the fine structure of these compounds, although the formation of complexes has been advanced as evidence for the 0x0 structure in some cases.5D Recently, the tautomerism of 5,5’-dihydroxydipyrromethanes and the corresponding 5-hydroxy-5‘-ethoxy derivatives has been investi“H. Fischer, T. Yoshioka, and P. Hartmann, Z. physiol. Chem. 212, 146 (1932). “T.S.Stevens, in “Chemistry of the Carbon Compounds” (E. H. Rodd, ed.), Vol. IVB, p. 1111 ff. Elsevier, Amsterdam, 1959. “ K . W. Bentley, “The Natural Pigments,” p. 162. Interscience, New York, 1960.
Mdh Hy 14
A. R. KATRITZKY AND J . M. LAGOWSKI
-
CH2'
H
M
0
Me
I
~
Et
A
Me
I
I
H
H
NL~
CHA ,CH&O& ~
~
Me
H ~ 3 1
[fjZl
gated.51a Two forms were isolated in each class which were assigned structures 63a (R = H, Et), 63b, and 63c on the basis of ultraviolet, infrared, and proton resonance spectra.
Meac& Me
0
Me
I
' H A
H
H
have discussed equilibria involving the Gray and his side chains of tetrapyrrole bile pigments.
F. P-HYDROXYPYRROLES Less is known about the P-hydroxypyrroles than about the isomeric a-hydroxy compounds. Originally ethyl 4-hydroxy-2-methylpyrrole-3carboxylate was suggested, on the basis of chemical evidence, to exist as a mixture of the 0x0 and hydroxy forms, 64 and 65,respectively.G0 OoaC.H. Gray, A. Kulczycka, and D. C. Nicholson, J . Chem. SOC. pp. 2268, 2276 (1961). Benary and B. Silbermann, Ber. deut. chem. Ges. 46, 1363 (1913).
TAUTOMEBISM: 111. &MEMBERED RINGS-1
HETERO ATOM
15
The chemical reactions of this compound were recently reconsidered, and both structures 64 and 65 were “rejected” in favor of the zwitterion formulation 66, which is supported by the presence of a band a t 3 . 1 ~(3226 cm-I) in the infrared spectrumG1and is merely an alternative canonical form of 64. On the other hand, the ultraviolet spectrum of 4-hydroxypyrrole-2-carboxylicacid (67) resembles that of its ethyl ether, possibly indicating that the 2-acid exists in the hydroxy form.6z
G.
P O L Y - 0 x 0 - AND -HYDROXY-PYRROLES AND -1NDOLES
Infrared and ultraviolet spectral data indicate that 1,5-diarylpyrrolidine-2,3-diones exist in the monooxo form Chemical evidence was advanced for a tautomeric equilibrium between 69 and 70,64and later spectroscopic work showed that 70 was the predominant form.65Compounds of type 71 were formulated, without experimental evidence, in the 0x0-imino form:* although the tendency for C=NR-+ C-NHR is usually greater than that for C-0 + C-OH in analogous cases. “ A . Treibs and A. Ohorodnik, Ann. Chem. Liebigs 611, 149 (1958). a’ R. Kuhn and G. Osswald, Chem. Ber. 89, 1423 (1956). W. R. Vaughan and L. R. Peters, J . Org. Chem. 18, 382 (1953). O’L. Horwitz, J. Am. Chem.. SOC.75, 4060 (1Y53). “D. G. O’Sullivan and P. W. Sadler, J. Chem. Soc. p. 876 (1959).
16
A. R. KATRITZKY AND J. M. LAGOWSKI
Ar
k
In 1882 Baeyer and OekonomidestiGadvanced formula 72 (R = H) for isatin on chemical grounds, but shortly thereafter the dioxo structure 73 (R = H) was proposed since the ultraviolet spectrum of isatin resembled that of the N-Me derivative (73, R = Me) and not that of the 0-Me derivative (72,R = Me).G7It was later shown, despite a conflicting report,Gsthat the ultraviolet spectrum of isatin is very similar to the spectra of both the 0- and N-Me derivat i v e ~; ~the~ early - ~ ~investigators had failed to take into consideration t,he facile decomposition of the 0-Me derivative. Although isolation of the separate tautomers of isatin has been reported,i2 these claims were disproved.'l A first attempt to determine the position of the mobile hydrogen atom using X-ray crystallographic techniques was i n c o n c l ~ s i v e ,but ~ ~ later X-ray dipole moment data,75 and especially the infrared spectrumiG demonstrated the correctness of the "A. v. Baeyer and S.Oekonomides, Ber. deict. chem. Ges. 15, 2093 (1882). "W. N. Hartley and J. J. Dobbie, J. Chem. Soc. 75, 640 (1899). "*J.Dabrowski and L. Marchlemski, Bull. SOC. chim. Fiance 53, 946 (1933). a*R.A. Morton and E. Rogers, J. Chem. Soc. 127, 2698 (1925). 'OR. G. Ault, E. L. Hirst, and R. A. Morton, J. Chem. Soc. p. 1653 (1935). " A . Hantzsch, Ber. deut. chem. Ges. 54, 1221 (1921). "G. Heller, Ber. deut. chem. Ges. 53, 1545 (1920). "E. G. Cox, T. H. Goodwin, and A. I. Wagstaff, Proc. Boy. Soc. A157, 399 (1936). IrG. H.Goldschmidt and F. J. Llewellp, Acto Cryst. 3, 294 (1950). "E. G. Cowley and J. R. Partington, J. Chem. Soc. p. 47 (1936). "D. G. O'Sullivan and P. W. Sadler, J. Chem. SOC.p . 2202 (1956).
TAUTOMERISM : 111. 5-MEMBERED RINGS-1
HETERO ATOM
17
dioxo formulation 73 (R = H). The polarographic behavior of isatin has also been discussed in relation to its tautomerism.77f78 7-Methylisatin-4-carboxylic acid has been reported to be in equilibrium with the tricyclic structure 74.79
Both the infrared and ultraviolet spectra of pyrrolidine-2,3,5-triones (75) have been interpreted to support their existence as hydroxymaleimides (76),80and the occurrence of a strong OH stretching band in the infrared spectrum of 4-phenylpyrrolidine-2,3,5-trione has been taken as evidence that it too exists in a hydroxy form, probably 76 (R = CsH,).*' However, the trioxo formulation is suggested by the infrared spectra of N-substituted pyrrolidine-2,3,Btriones, although an equilibrium apparently occurs depending upon the substituents and conditions.82 The zwitterion formulation 77 has been advanced for 4-amin0pyrrolidine-2,3,5-trione.~~ For chemical evidence
W. C. Sumpter, J. L. Williams, P. H. Wilkin, and B. L. Willoughby, J . O r g . Chem. 14, 713 (1949). '*W. C. Sumpter, P. H. Wilkin, J. L. Williams, R. Wedemeyer, F. L. Boyer, and W. W. Hunt, J . Org. Chem. 16, 1777 (1951). "P. W. Sadler, H. Mix, and H. W. Krause, J . Chem. SOC. p. 667 (1959). 'OR. H. Wiley and 5.C. Slaymaker, J. Am. Chem. SOC.80, 1385 (1958). "G. S. Skinner and C. B. Miller, J . Am. Chem. SOC.75, 977 (1953). "J. C. Sheehan and E. J. Corey, J. Am. Chem. SOC.74, 360 (1952). "E. G. Howard, A. Kotch, R. V. Lindsey, and R. E. Puttnam, 133rd Meeting, Am, Chem. SOC., Abstr., p. 65N (April 1958, San Francisco). "
18
A. R. KATRITZKY AND J. M. LAGOWSKI
concerning the tautomerism of these compounds, see reference 82 and references therein.
H. MONOHYDROXYINDOLES An initial study of the infrared spectrum of oxindole purported to show the presence both of the hydroxy (78) and of the 0x0 forms (79),S4 and, indeed, chemical evidence led to the same c o n c l u ~ i o n . ~ ~ On the basis of later infrared work, however, oxindole and its derivatives were considered to exist more or less completely in the 0x0 form,s6 and this conclusion is supported by ultraviolet spectroscopic data, i.e., by comparison of the spectrum of the parent compound with those of its methyl derivative^.^? The infrared spectrum of
ari
OH
I
H [781
I
Me [801
H [ 791
I
Me [811
1-methyloxindole-3-aldehyde is in accord with its formulation as 80 and/or 81.85 I n 1883, the hydroxy structure 82 was assigned to indoxyl on the basis of chemical evidence.88 More recently, however, the infrared spectra of 1-acetyl- and 1-methyl-indoxyl measured in chloroform indicated that the 0x0 form 83 (R = Ac, Me) greatly predominates,4°pso "E. D. Bergmann, J. A m . Chem. SOC. 77, 1549 (1955). " P . L. Julian, J. Pikl, and F. E. Wantz, J. A m . Chem. SOC.57, 2026 (1935). = A . E. Kellie, D. G . O'Sullivan, and P. W. Sadler, J. Chem. SOC.p. 3809 (1956). Mme. Ramart-Lucas and Mlle. Biquard, Bull. S O C . chim. France 2, 1383 (1935). 88 A. von Baeyer, Ber. deut. chem. Ges. 16, 2188 (1883). BeH.C. F. Su and K . C. Tsou, J. A m . Chem. SOC.82, 1187 (1960).
TAUTOMERISM: 111. &MEMBERED RINGS-1
HETERO ATOM
19
but the presence of an acetyl group in the 2-position has been shown to cause enolization to The important dyestuff-intermediate, indigo white, is usually formulated as 85 (see, e.g., reference go), but
Ho@ J
I
H
R
~4 1
apart from “general phenolic character” there is little definite evidence to support this formulation.
I. OTHER COMPOUNDS WITH POTENTIAL HYDROXYL GROUPS Infrared and ultraviolet spectral comparisons by Grob and Hoferel show that the position of the equilibrium 86 e 87 predominantly favors 87.
OPT.S. Stevens, in “Chemistry of the Carbon Compounds” (E. H. Rodd, ed.), Vol. IVB, pp. 1081, 1094. Elsevier, Amsterdam, 1959. O’C. A. Grob and B. Hofer, Helv. Chim. Actu 36, 847 (1953).
20
A. R. KATRITZKY AND J. M. LAGOWSKI
111. Compounds with Potential Mercapto Groups Early work on thiophenethiols has been summarized by Hartough19? although few conclusions were reached concerning their tautomerism. Caesar and Brantong3 concluded from infrared spectral data that 3-mercaptothiophene existed a t least partly in the thiol form but that the 3,4-dithiol existed completely as the dithione 88. Gronowitz and his associates have recently made a definitive investigation of a series of thiophene-2- and -3-monothiols using nuclear magnetic
r e ~ o n a n c e ~and ‘ ~ ~infrared ~ ~ ~ ~ s p e ~ t r o s c o p y .Their ~ ~ ~ ~results ~ show convincingly that all of these compounds exist predominantly in the thiol form, both as the pure liquid and in cyclohexane solution. Mercapto-pyrr~les~’and -indo1esg8 have been characterized, but little is known about their tautomerism.
IV. Compounds with Potential Amino Groups The amino form is usually much more favored in the equilibrium between amino and imino forms than is the hydroxy form in the corresponding keto-enol equilibrium. Grob and Utzingerg9suggest that in the case of a-amino- and a-hydroxy-pyrroles, structure 89 increases the mesomeric stabilization and thus offsets the loss of pyrrole resonance energy, but the increase due to structure 90 is not sufficient to offset this loss. Similar reasoning may apply to furans and
’’H .
D. Hartough, “Thiophene and Its Derivatives,” pp. 42S429. Interscience, New York, 1952. “P. D. Caesar and P. D. Branton, Ind. Eng. Chem. 44, 122 (1952). M R .A. Hoffman and S. Gronowitz, Arkiv Kemi 16, 515 (1961). 95R.A; Hoffman and S. Gronowite, Arkiv Kemi 16, 563 (1961). MS. Gronowitz, P. Moses, and A. B. Hijrnfeldt, Arkiv Kemi 17, 237 (1961). ‘“P.Pratesi, Atti accad. Lincei 16, 443 (1932); Chem. Abstr. 27, 2442 (1933). ”W. C. Sumpter and F. M. Miller, “Heterocyclic Compounds with Indole and Carbarole Systems,” pp. 53-54. Interscience, New York, 1954. w C .A. Grob and H. Utzinger, Helv. Chim. Acta 37, 1256 (1954).
TAUTOMERISM: 111. &MEMBERED RINGS-1
HETERO ATOM
21
A
H
~ 9 1
[go1
thiophenes. The limited experimental evidence available certainly demonstrates that amino compounds are more stable as such than their hydroxy analogs.
A. AMINOFURANS The substituted 3-aminofurans (91, R = H, Me) resinify in air, can be diazotized (structure 91), and are easily hydrolyzed (structure 92 ? ) . l o oThe infrared spectra of both 2- and 3-acetamidofuran show a strong NH stretching band indicating that these compounds do, indeed, exist in the acetamido form.lol C0,Et m 2
Me [gza]
The a-aminobenzofuran 92a exists in the amino form shown, as evidenced by infraredlOlaand proton resonance spectra.lOlb ImH. B. Stevenson and J. R. Johnson, J. Am. Chem. SOC. 59, 2525 (1937). '"R. Kuhn and G. Kriiger, Chem. Ber. 89, 1473 (1956). lolaJ. Derkosch and I. Specht, Monatsh. Chem. 92, 542 (1961). lolbA. R. Katritzky and J. Derkosch, Monntsh. Chem. 93, 541 (1962).
22
A. R. KATRITZKY AND J. M. LAGOWSKI
B. AMINOTHIOPHENES The tautomerism of 2- and 3-aminothiophenes was mentioned by Hartough in his review of thiophenes,lo2but the first definite evidence became available in 1961 when Hoffman and Gronowitz’* showed conclusively by nuclear magnetic resonance spectroscopy that these compounds both exist in the amino form. In agreement with this finding, 3-aminothiophene generally behaves as an aromatic amine.lo3
C. AMINOPYRROLES Aminopyrroles are usually formulated as such because they are quite strong bases and do not easily lose ammonia by h y d r o l y s i ~ . ~ ~ ~ , ~ ~ ~ I n agreement with the amino formulation, the infrared spectrum of the a-aminopyrrole 93 (R = H) contains a vNH, doublet and a band near 1660 cm-l corresponding to the NH, deformation frequency, and the infrared spectrum of the acetamino derivative is in agreement with the structure 93 (R = Ac) .88 However, a stable imino compound, probably with structure 94, has been isolated.56
I
CH,CH (OEt), ~ 3 1
I
\
H
CO,But
[941
The relative stability of 2,5-diaminopyrrole- (95) and 2,5-diiminopyrrolidine-type (96) structures for succimidines has not yet been c1arified,100-108 although ultraviolet .spectral data indicate that forms of type 97 are probably unimportant.loD BanfieldllO has discussed H. D. Hartough, “Thiophene and Its Derivatives,” pp. 228-240. Interscience, New York, 1952. Io3E.Campaigne and P. A. Monroe, J. A m . Chem. Soc. 76, 2447 (1954). l‘C. A. Grob and P. Ankli, Helv. Chim. Acta 33, 273 (1950). losH. Fischer and H. Orth, “Die Chemie des Pyrrols,” Vol. 1, p. 110 ff. Akademische Verlag., Leipzig, 1934. lWG.E. Ficken and R. P. Linstead, J. Chem. Soc. p. 3525 (1955). ’O‘J. Schurz, A. Ullrich, and H. Bayzer, Monatsh. Chem. 90, 29 (1959). l m J . A. Elvidge, Chem. Sac. (London), Spec. Publ. N o . 4, p. 28 (1956). lO9 J. A. Elvidge and R. P. Linstead, J. Chem. Sac. p. 442 (1951). llOJ. E. Banfield, J. Chem. SOC.p. 2098 (1961).
TAUTOMERISM: 111. &MEMBERED RINGS-1 HETERO ATOM
H,N
rn N I H
HN QNH I
NH,
HN
H
23
NHZ
HZN. H,N
"H
/C=N [991
the fine structure of guanidino derivatives of type 99.
D. AMINOINDOLES Z-Aminoindole was initially assigned structure 100 (R = H) on chemical grounds"' and later structure 101 was erroneously assigned because the ultraviolet spectrum of its hydrochloride was similar to that of oxindole.llz In 1956, Kebrle and Hoffmannlls established the
Me
"'R. Pschorr and G. Hoppe, Ber. deu.t. chem. Ges. 43, 2543 (1910). "'H. Rinderknecht, H. Koechlin, and C . NiemanQ, J. Org. Chem. 18, 971 (1953). J. Kebrle and K. Hoffrnann, Helv. Chim. Acta 39, 116 (1956).
24
A. R. KATRITZKY AND J . M. LAGOWSKI
predominance of structure 102. 2-Amino-l-methylindole (100, R = Me) exists largely in the amino form in ethanol (see later), since its ultraviolet spectrum differs from that of 103 (R = Me). The ultraviolet spectrum of 2-aminoindole differs from both those of 100 (R = Me) and of 103 (R = Me) ; therefore, i t probably exists as 102. All three bases show similar ultraviolet spectra in acidic solutions indicating that they form similar cations, i.e., of type 104, and therefore the pK, method could be used to study these equilibria. Under all conditions, form 102 was found to be the most stable tautomer by a factor >lo. The equilibrium between 103 (R = H) and 100 (R = Me) is displaced toward 103 (R = H), the displacement increasing with increasing polarity of the solvent; a comparison of the ultraviolet spectra led to the same conclusion. Infrared spectra support these structural assignments. The acylamino compounds 105 ( R = H, Me, Ac)l13 and the isoindole derivative 106lI4 also exist in the forms shown on the basis of infrared and ultraviolet spectra1 evidence.
E. POTENTIAL AMINODERIVATIVES OF ISOINDOLE Ultraviolet spectral comparisons indicate that structure 107 predominates over 108 when R = H or OH, but that 107 is the predominant form when R = ~ r y 1 . lSimilarly, ~~ 109 predominates over 110 by a large factor when R = H , OH, or Me, and by a smaller factor when R is a higher alkyl group, but 110 predominates when R
is an aryl group. (For a discussion of guanidino derivatives corresponding to 110, see reference 110.)
V. Compounds with Potential Methyl Groups 2-Methylene-2,5-dihydrofuran (111) , which has been isolated by Rice,116is converted by acid into 2-methylfuran.
I w M . E . Baguley and J. A. Elvidge, J . Chem. Soc. p. 709 (1957). lX5P.F. Clark, J. A. Elvidge, and J. H. Golden, J . Chem. SOC.p. 4135 (1956). IlaH. L. Rice, J . Am. Chem. Soc. 74, 3193 (1952).
TAUTOMERISM : 111. &MEMBERED RINGS-1
HETERO ATOM
8
0
-
25
NHR
c
RN
VI. Other Substituted Pyrroles A. VINYLPYRROLES The tautomeric equilibrium between 112 and 113 has been studied by Treibs and his associate^.^^^^'^^ Most pyrromethenes take form
113 unless both rings carry electron-withdrawing substituents in which case structure 112 is predominant. Common cations of type 114 are favored by both tautomers.
Treibs and F. Reitsam, Ann. Chem. Liebigs 611, 194 (1958). ,"A. Treibs, E. Herrniann, E. Meissner, and A. Kuhn, Ann. Chem. Liebigs 602, InA.
153 (1957).
26
A. R. KATRITZKY AND J . M. LAGOWSKI
B. NITROSOPYRROLES Chemical evidence has been advanced for the formulation of p nitrosoindoles as either 115 or 116 (see reference 119 and references therein), but ultraviolet spectral comparison with both methylated forms clearly indicated that 116 was favored. Infrared spectral data are also considered to support structure 116.'*O
H [115J ~~
[lltij
~
I*'N. Campbell and R. C. Cooper, J. Chem. SOC.p. 1208 (1935). I2'F. Piozzi and M. Dubini, Gazz. chirn. ital. 89, 638 (1959).
. .
A R KATRITZKY? University Chemical Laboratory. Cambridge. England AND
. .
J M LAGOWSKI Genetics Foundation. The U i h e r s i f y of Texas. Artstin. Texas
I . Tautomerism of Pyrroles Not Involving the Functional Group I1. Compounds with a Potential Hydroxyl Group . . . . . A . Hydroxyfurans . . . . . . . . . . . . . . B. Potential Dihydroxyfurans . . . . . . . . . . . C . a-Hydroxythiophenes . . . . . . . . . . . . D . P-Hydroxythiophenes . . . . . . . . . . . . E . a-Hydroxypyrroles . . . . . . . . . . . . . F. p-Hydroxypyrroles . . . . . . . . . . . . . . G. Poly-0x0- and -hydroxy-pyrroles and -indoles . . . . . H . Monohydroxyindoles . . . . . . . . . . . . I . Other Compounds with Potential Hydroxyl Groups . . . I11. Compounds with Potential Mercapto Groups . . . . . . IV . Compounds with Potential Amino Groups . . . . . . . A . Aminofurans . . . . . . . . . . . . . . . B . Aminothiophenes . . . . . . . . . . . . . . C . Aminopyrroles . . . . . . . . . . . . . . . D . Aminoindoles . . . . . . . . . . . . . . . E . Potential Amino Derivatives of Isoindole . . . . . . V . Compounds with Potential Methyl Groups . . . . . . V I . Other Substituted Pyrroles . . . . . . . . . . . . A . Vinylpyrroles . . . . . . . . . . . . . . . B . Nitrosopyrroles . . . . . . . . . . . . . . Errata . . . . . . . . . . . . . . . . . .
.
3 . 5 . 5 . 6 . 8 . 9 . 1 1 . 1 4 . 15 . 1 8 . 19 . 2 0 . 20 . 2 1 . 2 2 . 2 2 . 2 3 . 2 4 . 24 . 25 . 25 . 2 6 . 26
The most important potentially tautomeric thiophenes and furans are those carrying hydroxyl. mercapto. and amino groups. I n these compounds a prototropic shift can occur between the functional group *The first two chapters in this volume conclude the series of four articles on prototropic tautomerism; the first two articles appeared in Volume 1. Cross references to these articles include. for easy identification. the roman numeral given in the title . t Present address : School of Chemistry. University of East Anglia. Norwich. England . 1
2
A. R. KATRITZKY AND J. M. LAGOWSKI
and a ring carbon atom, cf. type ii described in Volume 1, Section I,A, of article I by Katritzky and Lagowski. A single functional group in the a-position gives rise to three possible tautomeric structures, 1-3, whereas two structures, 4 and 5, are possible with one functional group in the P-position. Similar structures are possible for the analogous pyrroles, but pyrroles which do not carry a substituent on the hetero nitrogen atom can undergo additional types of tautomerism (see later). The proportion of the various forms present a t equilibrium is dependent upon their relative stability. Forms 1 and 4 are “aromatic,” and, since the aromatic stabilization decreases in the order thiophene > pyrrole > furan, forms of this type would be expected to be the most favored for thiophenes. The nonaromatic forms are, however, also mesomeric as indicated by curly arrows in 2, 3, and 5.
“1
Pl
[31
Systems 2 and 3 are cross-conjugated, but 5 is not, and i t might have been expected that a-substituted compounds would be more prone to exist in the aromatic form 1 than the ,&compounds to exist as 4. From the limited evidence available, the reverse appears to be the case. When the hetero atom is not a very strong electron donor, i.e., 2 = S or 0, structure 3 would be expected to be relatively more stable than 2, and this is supported by the evidence available. When 2 = NR, the difference in the stabilities of 2 and 3 could be smaller. The nature of the substituent group X plays an important role in determining the relative stability of the various tautomeric forms. I n aliphatic systems the tendency of :C=C-XH to become :CH-C=X increases markedly in the order X = CH, < NH < 0, and certainly hydroxy compounds show less inclination to exist as such than do amino compounds. The position of S in this series is not completely
TAUTOMERISM: 111. &MEMBERED RINGS-1
HETERO ATOM
3
clear (see discussion in Section I11 of article 11, Volume 1 ) ; however, the thione-thiol equilibrium 6 + 7 appears to be more in favor of 7 than in the case of the corresponding keto-enol equilibrium.’ In agreement with this contention, mercaptothiophenes are more stable than the hydroxythiophenes (discussed later). Electron-withdrawing substituents stabilize the aromatic forms 1 and 4, as would be expected, by acting as an “electron sink” and by forming intramolecular hydrogen bonds with the XH group when this is sterically possible. On the other hand, a benzene ring fused in the 3,4-position greatly stabilizes 8 as compared to 9. The effect of a benzene ring fused in the 4,5-position is smaller, but probably also causes some preferential stabilization of the nonaromatic forms corresponding to 2 and 5. R-C-CCH,CO,Et
II S
=R-CCCH-C0,Et I SH
I. Tautomerism of Pyrroles Not Involving the Functional Group The tautomeric forms 11 and 12, called pyrrolenines, have often been postulated for pyrrole (lo), but there is no conclusive evidence for their existence.2 The report3 that two isomers exist as the pyrrolenine forms 13 and 14 (Ar = 3,4-dimethoxyphenyl) must be regarded with considerable doubt. Chemical evidence led to the conclusion that the conjugate acids of pyrrole probably exist predominantly as 16 or 17 rather than as 15.2,4,5Although infrared spectra were initially interpreted on the ‘ Z . Reyes and R. M. Silverstein, J . Am. Chem. SOC.80, 6367 (1958). ‘ A . Treibs and G. Fritz, Ann. Chem. Liebigs 611, 162 (1958). *D. A. Guthrie, A. W. Frank, and C. B. Purves, Can. J . Chem. 33, 729 (1955). ‘A.Treibs and K . H. Michl, Ann. Chem. Liebigs 577, 129 (1952). ‘A. Treibs and A. Ohorodnik, Ann. Chem. Liebigs 611, 139 (1958).
4
A. R. KATRITZKY AND J . M. LAGOWSKI
H
1131
~ 4 1
basis of structure 15: nuclear magnetic resonance spectra later proved that structures of type 16 were favored in ~ o l u t i o n Similarly, .~ ultraviolets and nuclear magnetic resonance spectras and basicity measurem e n t ~show ~ ~ that indole cations are formed by protonation a t the
p-position to give 18, substantiating the postulation which had been offered earlier on theoretical g r o ~ n d s . ~ Dipyrromethenes rapidly exchange the hydrogen atom attached to nitrogen, and the two isomers of unsymmetrical compounds (e.g., 19 and 20) cannot be separately isolated.lDDipyrromethenes form meso-
'E. Bullock, Can. J . Chem. 36, 1686 (1958). 'R. J. Abraham, E. Bullock, and S. S. Mitra, Can. J . Chem. 37, 1859 (1959). * M. J. Kamlet and J. C. Dacons, J . Org. Chem. 26, 220 (1961). OR.L. Hinnar and J. Lang, Tetrahedron, Letters No. 21, 12 (1960).
"G.Berti, A. Da Sett,imo, and D. Segnini, Gnzz. chim. ital.
91, 571 (1961).
1°H. Fischer and B. Walach, Ann. Chem. Liebigs 450, 109 (1926).
TAUTOMERISM : 111. &MEMBERED RINGS-1
n
HETERO ATOM
5
n
I
P I
meric cations (cf. 21) by proton addition a t n i t r ~ g e n whereas ,~ olefinic derivatives such as 22 undergo proton addition on carbon. 11. Compounds with a Potential Hydroxyl Group Heterocyclic compounds carrying hydroxyl groups may be compared with phenols. Thomsonll has reviewed the tautomeric behavior of phenols; often both tautomeric forms of polycyclic compounds such as naphthols can be isolated. Early work on hydroxythiophenes and -furans was also reviewed by Thomson,'l but until recently their chemistry has been in a somewhat confused state. A pattern is now beginning to emerge, a t least for the a-substituted con~pounds,which appear to exist as a3-ox0 derivatives and to attain equilibrium slowly with the corresponding a4-0x0 forms. For the a-hydroxy compounds, the equilibrium generally favors the a3-ox0 form.
A. HYDROXYFURANS Unsaturated 7-lactones, e.g., a- (23) and ,&angelica lactone (24), are well known. Compounds 23 and 24 are both converted by alkaline catalysts into an equilibrium mixture in which 23 predominates, the amount of the hydroxy form (25) present a t equilibrium being exceedingly small. True a-hydroxyfurans are unknown, and, although the preparation of both a- and ,f3-hydroxyfurans has been rep0rted,1~J~ these claims have often been refuted (see, e.g., reference 14). Me a 0
Me
Me
EL
"R. H. Thomson, Quart. Revs. (London) LO, 27 (1956). "H. H. Hodgson and R. R. Davies, J. Chem. SOC.p. 806 (1939). "H. H. Hodgson and R. R. Davies, J . Chem. SOC.p. 1013 (1939). I'M. P. Cava, C. L. Wilson, and C. J. Williams, J . A m . Chem. Soc. 78, 2303 (1956).
6
A. R. KATRITZKY AND
J. M. LAGOWSKI
Very little is known concerning the simple, monocyclic 3-hydroxyfurans (cf. reference 15). Both the 0x0 and hydroxy forms of the substituted 3-hydroxyfurans 26 and 27 (R = H, C,H,) have been isolated>G,li but the individual tautomers slowly undergo interconversion. The enol forms give a positive reaction with ferric chloride, react rapidly with bromine, and form a peroxide with oxygen. From chemical evidence, the benzo derivatives of 3-hydroxyfuran, 28lS and 29,19appear to exist predominantly in the 0x0 form, and this is further supported by ultraviolet spectral data.20 Stefanye and Howard?’
P I
~ 9 1
concluded from infrared spectroscopic data that 5,7-dichloro-3-hydroxybenzofuran exists in the 0x0 form, but that the hydroxy form of its 2- (5’,7’-dichlorobenzofuryl) derivative, i.e., 3-hydroxy-5,5’,7,7’tetrachloro-2,3’-bibenzofuran, apparently predominates.
B.
POTENTIhL
DIHTDROXYFURANS
a-Keto-y-lactones appear to exist in the dioxo form 30,22 but enolization can occur when this leads to extended conjugation. The I5E. VotoEek and S. Malachta, Collection Czechoslov. Chem. Communs. 4, 87 ( 1932). leE. P. Kohler, F. H. Westheimer, and M. Tishler, J . A m . Chem. Sac. 58, 264 (1936). “ E . P.Kohler and D. W. Woodward, J. A m . Chem. SOC.58, 1933 (1936). “ K . v. Auwers and E. Auffenberg, Ber. deut. chem. Ges. 52, 92 (1919). ”P. Emmott and R. Livingstone, J. Chem. SOC.p. 4629 (1958). *Mme. Ramart-Lucas and M. van Cowenbergh, Bull. sac. chim. France p. 1381 (1935). ‘ID. Stefanye and W. L. Howard, J . 01.8.Chem. 20, 813 (1955). P1. A. Plattner and L. M. Jampolsky, Helv. Chim. Acta 26, 687 (1943).
TAUTOMERISM : 111. 5-MEMBERED mNGs-1
HETERO ATOM
7
enol forms of 31 (R = C,H,) and 31 (R = H) have been shown to be predominant by chemical evidencez3 and by infrared spectral datalZ4
~301
[311
respectively. Tetronic acids exist predominantly in the dioxo form (32) in solvents of low polarity, while the existence of the monoenol form (33) has been established in other solvents by infraredz5 and ultraviolet spectral comparisonsz6 and from dipole moment dat.a.27Haynes and Plimmerz7ahave recently reviewed the structure of these compounds [see also reference 28(a)], and the tautomerism of vitamin A (U), which has a related structure, has also been surveyed.z8(b)Analogous compounds carrying an amino group in the 3-position are also known.28(c)
Succinic anhydride (35, Z = 0) can theoretically tautomerize to 36, but all the evidence indicates that i t exists overwhelmingly as 35; for example, the infrared spectrum shows v C=O 23W. E. Bachmann, G. I. Fujimoto, and L. B. Wick, J. Am. Chem. SOC.72, 1995 (1950).
*’L. Mangoni and M . Belardini, Ann. chim. (Rome) 50, 322 (1960). “ L . A. Duncanson, J. Chem. SOC.p. 1207 (1953). “ E . R. H. Jones and M. C. Whiting, J. Chem. Soc. p. 1419 (1949). ” W . D. Kumler, J. Am. Chem. Soc. 62, 3292 (1940). *“L. J. Haynes and J. R. Plimmer, Quart. Revs. (London) 14, 292 (1960). “H. von Euler and B. Eistert, “Chemie und Biochemie der Reduktone und Reduktonate,” (a) p. 159, (b) p. 185, (c) p. 261. F. Enke, Stuttgart, Germany, 1957.
H. M. Randall, R. G. Fowler, N. Fuson, and J. R. Dangl, “Infrared Determination of Organic Structures.’’ D. Van Nostrand, New York, 1949.
8
A. R. RATRITZKY AND J . M. LAGOWSKI
Infrared and nuclear magnetic resonance spectral evidence led Kendall and Hajos3" to conclude that furan-3,4-dione (37) exists as such, which is surprising since cyclic a-diketones with five-membered rings are usually monoenolized. The 2,bdicarbethoxy derivative 38 was earlier stated to exist in the dihydroxy form.31 0a 0H
O
OOH
,OH
H\ O
V
[371
0
CO,E t
Et0,C [381
C. a-HYDROXYTHIOPHENES The infrared spectrum of 2-hydroxythiophene was originally interpreted as showing both v O H and vC=O peaks indicating that it exists as a mixture of the hydroxy form 39 (R = H ) and a t least one of the 0x0 forms, 40 and/or 41.32The ultraviolet spectrum of 2-hydroxythiophene is different from that of the corresponding methyl ether (39, R = Me) suggesting the presence of the chromophore contained in structure 40. The facts that this compound gives a positive
color test with ferric chloride, is a weak acid, and undergoes reactions characteristic of a phenolic hydroxyl group have been advanced as further evidence for the presence of the hydroxy form.32The infrared % E .C . Kendall and 2. G. Hajos, J. Am. Chem. Soc. 82, 3219 (1960). 3i W.H. Hoehn, Iowa State Coll. .I. Sci. 11, 66 (1936); Chem. Abstr. 31, 1800 (1937). "C.D.Hurd and K. L. Kreuz, J. Am. Chem. SOC.72, 5543 (1950).
TAUTOMERISM : 111. &MEMBERED RINGS-1
HETERO ATOM
9
spectrum of 5-phenyl-2-hydroxythiophene measured in chloroform solution shows a C=O absorption band, but its ultraviolet spectrum in ethanol is rather similar to that of the corresponding 2-methoxy derivative. Comparison of the ultraviolet spectra was complicated by a spontaneous oxidation of the latter compound. The ultraviolet spectrum measured in chloroform or isooctane did not correspond to that obtained in ethanol. Chemical evidence was also considered to support a mobile keto-enol tautomeric e q ~ i l i b r i u m . ~ ~ The foregoing conclusions must now be modified on the basis of a recent, detailed investigation of the tautomerism of 2-hydroxy5-methylthiophene by Gronowite and Hoffman3* using nuclear magnetic resonance and infrared spectroscopy. Compounds 42 and 43 form an equilibrium mixture containing 85% of 43. However, equilibrium is attained slowly, and 42 can be obtained essentially pure by dissolving the mixture in a base and precipitating with acid, whereas almost pure 43 can be isolated by distillation. The effect of the methyl group in stabilizing structure 42 is illustrated by the fact that 2hydroxythiophene itself exists essentially completely as 40.34
and it 2-Hydroxythianaphthene has been isolated in two has been suggested that these may be the individual keto and enol tautomers.ll A reinvestigation of this system using modern techniques would be welcome.
D. /3-HYDROXYTHIOPHENES The infrared spectrum of 3-hydroxythiophene has been interpreted by Ford and M a ~ k a to y ~show ~ that it exists as a mixture of both the hydroxy (44) and the 0x0 forms (45). 5-Phenyl-3-hydroxythiophene apparently behaves similarly to the 2-hydroxy isomer, the ultraviolet asA. I. Kosak, R. J. F. Palchak, W. A. Steele, and C. M. Selwitz, J . Am. Chem. SOC.76, 4450 (1954). Gronowitz and R. A . Hoffman, Arkiv Kemi 15, 499 (1960). " C. Marschalk, J . prakt. Chem. 88, 227 (1913). "M. C. Ford and D. Mackay, J . Chem. SOC.p. 4985 (1956).
10
A. R. KATRITZKY AND J. M. LAGOWSKI
spectrum measured in ethanol suggesting that it exists in the hydroxy form, whereas that determined in chloroform indicates that one of the 0x0 forms predon~inates.~~ An alkoxycarbonyl group in the 2position of the thiophene nucleus probably stabilizes a hydroxyl group in the 3-position (cf. reference 37). I n 1920, Auwers and T h i e ~ suggested ,~ that 3-hydroxythianaphthene
might exist as 46 in the solid &ate, but partly as 47 in alcoholic solution. These conclusions appear to be supported in part by more recent although infrared spectral data indicate that the 0x0
form (46) greatly predominates in both chloroform and carbon disulfide solutions.4o Snyder and his c o - ~ o r k e r s ~ ' assigned ~*~ structures 48 and 49 to these p-hydroxythiophene derivatives on the basis of chemical evidence and infrared and nuclear magnetic resonance spectral data. Infrared and nuclear magnetic resonance spectra further indicate that compounds of type 49 exist as dimers, probably hydrogen bonded, when R = OC,H, or CH,, but as monomeric enols when R = H.', arH.Fiesselmann and P. Schipprak, Chem. Ber. 89, 1897 (1956). 'K. v. Auwers and W. Thies, Ber. deut. chem. Ges. 53, 2285 (1920). *'F. Krohnke, Chem. Ber. 92, cxiv (1959). J. Holt, A. E. Kellie, D. G. O'Sullivan, and P. W. Sadler, J. Chem. SOC. p. 1217 (1958).
W. Carpenter and H. R. Snyder, J. Am. Chem. SOC.82, 2592 (1960). '*D. S. Matteson and H. R. Snyder, J . Org. Chem. 22, 1500 (1957). r 3 R .J. Tuite, A. D. Josey, and H. R. Snyder, J. Am. Chem. Soc. 82, 4360
"
( 1960).
TAUTOMERISM : 111. &MEMBERED
RINGS-1
HETERO ATOM
11
On the basis of chemical evidence, Swiss investigators have postulated that 3,4-dihydroxythiophenes exist as diols, i.e., as 50.44145 H
H
E. a-HYDROXYPYRROLES 2-Hydroxypyrroles are thought to exist in 0x0 forms such as 51 or 52; structure 53 illustrates a third possible 0x0 form.46 Chemical evidence for tautomerism in the hydroxypyrroles has been reviewed by Fischer and Orth.47 Since 2-hydroxypyrrole itself is unstable and Q-0I
H [511
0
I H [521
0
M?AyA HCI CO,R
0
[531
rapidly resinifies, studies have been confined to some of its more stable derivatives. The ultraviolet spectrum of 54 (R = C0,Et) is different from that of 55, and 54 (R = C0,Et) does not give a positive test with ferric ~ h l o r i d e , ~which ~ , ~ ~led to its formulation as shown. On the basis of their infrared (and ultraviolet) spectra, compounds 56 (R = H),5O 56 (R = A c ) , ~ "and 5751 must exist in 0x0 forms since they exhibit v C=O absorption bands: 56 (R = H) a t 5 . 9 5 ~(1681 cm-I), 57 a t 5 . 9 0 ~(1695 cm-l), and 56 (R = Ac) two bands a t 5.87 and 5 . 9 5 ~(1704 and 1681 cm-'). Proton resonance " P . Karrer and F. Kehrer, Helv. Chim. Acta 27, 142 (1944). "P. Karrer, R. Keller, and E. Usteri, Helv. Chim. Acta 27, 237 (1944). "H. Lapin and A. Horeau, Bull. SOC. chim. France p. 1703 (1960). " H.Fischer and H. Orth, "Die Chemie des Pyrrols." Vol. I, p. 124.Akademische Verlag., Leipzig, 1934. "C. A. Grob and P. Ankli, Helv. Chim. Acta 32, 2010 (1949). "C.A. Grob and P. Ankli, Helv. Chim. Acta 32, 2023 (1949). 'OH.Plieninger and M . Decker, Ann. Chem. Liebigs 598, 198 (1956). 61K. E. Schulte, J. Reisch, and R. Hobl, Arch. Pharm. 293, 687 (1960).
12
A. R. KATRITZKY AND J. M. LAGOWSKI
bh
R
spectra confirm structure 56 (R = H).51aThe A4-structure was assigned to 57 because it added water reversibly, and this is considered to be a characteristic reaction of A4-pyrrolones,52whereas the A3structure was assigned to 56 (R = H) because the ultraviolet spectrum appeared to be of the crotonic acid type.5o Ultraviolet spectra suggested, and nuclear magnetic resonance spectra proved, that 58 (R = H, OH) should be assigned the A3-structure The tautoineric behavior of compounds of type 59 has been discussed by Meyer and V a ~ g h a n .An ~ ~ intramolecularly hydrogenbonded 0x0 structure has been assigned to 60 on the basis of its infrared spectrum;5 whereas unambiguous chemical evidence, i.e., ozonolysis to succimide, confirmed the isolation of 61 in the 0x0 form.56 The foregoing results may be summarized as follows: potential a-hydroxypyrroles exist as pyrrolones. Substituents in the 3-position and in the 5-position favor the A3- and the A4-pyrrolone structure, respectively, as is to be expected. For 3,4,5-trisubstituted compounds, such as 61b, the A3-structure appears to be preferred.51aAn electron&'OH. Plieninger, H. Bauer, and A. R. Katritzky, Ann. Chem. Liebigs 654, 165 ( 1962).
"K,E. Schulte and J. Reisch, Arch. Phnrm. 292, 51 (1959).
53J. A. Moore and J. Binkert, J . A m . Chem. SOC.81, 6029 (1959). "W. L. Meyer and W. R . Vaughan, J. Org. Chem. 22, 1565 (1957). "P. L. Southwick and R. J. Owellen, J. Org. Chem. 25, 1133 (1960). uJ. A. Elvidge, J. S. Fitt, and R. P. Linstead, J. Chem. SOC.p. 235 (1956).
TAUTOMERISM : 111. &MEMBERED RINGS-1
HETERO ATOM
13
0
H
-NR
accepting substituent in the 4-position favors the A4-pyrrolone structure, because conjugation between the substituent and the cyclic nitrogen atom is then possible. An electron-accepting substituent in the 5-position might be expected to stabilize the hydroxypyrrole form, because only then is extended conjugation possible in the molecule: the aldehyde 61a does, indeed, exist as shown on proton resonance spectral
Me)rr7Me
OHC
N
Me\
,Me
OH
Interesting tautomeric possibilities exist in the xanthobilirubic acid series (cf. reference 57) which can be illustrated by the equilibrium 62$63. More complex examples of the same type are found among the linear tetrapyrrole pigments-the bilenes, bilidienes, and bilitrienes-and have been discussed by Stevens.58 Relatively little evidence is available concerning the fine structure of these compounds, although the formation of complexes has been advanced as evidence for the 0x0 structure in some cases.5D Recently, the tautomerism of 5,5’-dihydroxydipyrromethanes and the corresponding 5-hydroxy-5‘-ethoxy derivatives has been investi“H. Fischer, T. Yoshioka, and P. Hartmann, Z. physiol. Chem. 212, 146 (1932). “T.S.Stevens, in “Chemistry of the Carbon Compounds” (E. H. Rodd, ed.), Vol. IVB, p. 1111 ff. Elsevier, Amsterdam, 1959. “ K . W. Bentley, “The Natural Pigments,” p. 162. Interscience, New York, 1960.
Mdh Hy 14
A. R. KATRITZKY AND J . M. LAGOWSKI
-
CH2'
H
M
0
Me
I
~
Et
A
Me
I
I
H
H
NL~
CHA ,CH&O& ~
~
Me
H ~ 3 1
[fjZl
gated.51a Two forms were isolated in each class which were assigned structures 63a (R = H, Et), 63b, and 63c on the basis of ultraviolet, infrared, and proton resonance spectra.
Meac& Me
0
Me
I
' H A
H
H
have discussed equilibria involving the Gray and his side chains of tetrapyrrole bile pigments.
F. P-HYDROXYPYRROLES Less is known about the P-hydroxypyrroles than about the isomeric a-hydroxy compounds. Originally ethyl 4-hydroxy-2-methylpyrrole-3carboxylate was suggested, on the basis of chemical evidence, to exist as a mixture of the 0x0 and hydroxy forms, 64 and 65,respectively.G0 OoaC.H. Gray, A. Kulczycka, and D. C. Nicholson, J . Chem. SOC. pp. 2268, 2276 (1961). Benary and B. Silbermann, Ber. deut. chem. Ges. 46, 1363 (1913).
TAUTOMEBISM: 111. &MEMBERED RINGS-1
HETERO ATOM
15
The chemical reactions of this compound were recently reconsidered, and both structures 64 and 65 were “rejected” in favor of the zwitterion formulation 66, which is supported by the presence of a band a t 3 . 1 ~(3226 cm-I) in the infrared spectrumG1and is merely an alternative canonical form of 64. On the other hand, the ultraviolet spectrum of 4-hydroxypyrrole-2-carboxylicacid (67) resembles that of its ethyl ether, possibly indicating that the 2-acid exists in the hydroxy form.6z
G.
P O L Y - 0 x 0 - AND -HYDROXY-PYRROLES AND -1NDOLES
Infrared and ultraviolet spectral data indicate that 1,5-diarylpyrrolidine-2,3-diones exist in the monooxo form Chemical evidence was advanced for a tautomeric equilibrium between 69 and 70,64and later spectroscopic work showed that 70 was the predominant form.65Compounds of type 71 were formulated, without experimental evidence, in the 0x0-imino form:* although the tendency for C=NR-+ C-NHR is usually greater than that for C-0 + C-OH in analogous cases. “ A . Treibs and A. Ohorodnik, Ann. Chem. Liebigs 611, 149 (1958). a’ R. Kuhn and G. Osswald, Chem. Ber. 89, 1423 (1956). W. R. Vaughan and L. R. Peters, J . Org. Chem. 18, 382 (1953). O’L. Horwitz, J. Am. Chem.. SOC.75, 4060 (1Y53). “D. G. O’Sullivan and P. W. Sadler, J. Chem. Soc. p. 876 (1959).
16
A. R. KATRITZKY AND J. M. LAGOWSKI
Ar
k
In 1882 Baeyer and OekonomidestiGadvanced formula 72 (R = H) for isatin on chemical grounds, but shortly thereafter the dioxo structure 73 (R = H) was proposed since the ultraviolet spectrum of isatin resembled that of the N-Me derivative (73, R = Me) and not that of the 0-Me derivative (72,R = Me).G7It was later shown, despite a conflicting report,Gsthat the ultraviolet spectrum of isatin is very similar to the spectra of both the 0- and N-Me derivat i v e ~; ~the~ early - ~ ~investigators had failed to take into consideration t,he facile decomposition of the 0-Me derivative. Although isolation of the separate tautomers of isatin has been reported,i2 these claims were disproved.'l A first attempt to determine the position of the mobile hydrogen atom using X-ray crystallographic techniques was i n c o n c l ~ s i v e ,but ~ ~ later X-ray dipole moment data,75 and especially the infrared spectrumiG demonstrated the correctness of the "A. v. Baeyer and S.Oekonomides, Ber. deict. chem. Ges. 15, 2093 (1882). "W. N. Hartley and J. J. Dobbie, J. Chem. Soc. 75, 640 (1899). "*J.Dabrowski and L. Marchlemski, Bull. SOC. chim. Fiance 53, 946 (1933). a*R.A. Morton and E. Rogers, J. Chem. Soc. 127, 2698 (1925). 'OR. G. Ault, E. L. Hirst, and R. A. Morton, J. Chem. Soc. p. 1653 (1935). " A . Hantzsch, Ber. deut. chem. Ges. 54, 1221 (1921). "G. Heller, Ber. deut. chem. Ges. 53, 1545 (1920). "E. G. Cox, T. H. Goodwin, and A. I. Wagstaff, Proc. Boy. Soc. A157, 399 (1936). IrG. H.Goldschmidt and F. J. Llewellp, Acto Cryst. 3, 294 (1950). "E. G. Cowley and J. R. Partington, J. Chem. Soc. p. 47 (1936). "D. G. O'Sullivan and P. W. Sadler, J. Chem. SOC.p . 2202 (1956).
TAUTOMERISM : 111. 5-MEMBERED RINGS-1
HETERO ATOM
17
dioxo formulation 73 (R = H). The polarographic behavior of isatin has also been discussed in relation to its tautomerism.77f78 7-Methylisatin-4-carboxylic acid has been reported to be in equilibrium with the tricyclic structure 74.79
Both the infrared and ultraviolet spectra of pyrrolidine-2,3,5-triones (75) have been interpreted to support their existence as hydroxymaleimides (76),80and the occurrence of a strong OH stretching band in the infrared spectrum of 4-phenylpyrrolidine-2,3,5-trione has been taken as evidence that it too exists in a hydroxy form, probably 76 (R = CsH,).*' However, the trioxo formulation is suggested by the infrared spectra of N-substituted pyrrolidine-2,3,Btriones, although an equilibrium apparently occurs depending upon the substituents and conditions.82 The zwitterion formulation 77 has been advanced for 4-amin0pyrrolidine-2,3,5-trione.~~ For chemical evidence
W. C. Sumpter, J. L. Williams, P. H. Wilkin, and B. L. Willoughby, J . O r g . Chem. 14, 713 (1949). '*W. C. Sumpter, P. H. Wilkin, J. L. Williams, R. Wedemeyer, F. L. Boyer, and W. W. Hunt, J . Org. Chem. 16, 1777 (1951). "P. W. Sadler, H. Mix, and H. W. Krause, J . Chem. SOC. p. 667 (1959). 'OR. H. Wiley and 5.C. Slaymaker, J. Am. Chem. SOC.80, 1385 (1958). "G. S. Skinner and C. B. Miller, J . Am. Chem. SOC.75, 977 (1953). "J. C. Sheehan and E. J. Corey, J. Am. Chem. SOC.74, 360 (1952). "E. G. Howard, A. Kotch, R. V. Lindsey, and R. E. Puttnam, 133rd Meeting, Am, Chem. SOC., Abstr., p. 65N (April 1958, San Francisco). "
18
A. R. KATRITZKY AND J. M. LAGOWSKI
concerning the tautomerism of these compounds, see reference 82 and references therein.
H. MONOHYDROXYINDOLES An initial study of the infrared spectrum of oxindole purported to show the presence both of the hydroxy (78) and of the 0x0 forms (79),S4 and, indeed, chemical evidence led to the same c o n c l u ~ i o n . ~ ~ On the basis of later infrared work, however, oxindole and its derivatives were considered to exist more or less completely in the 0x0 form,s6 and this conclusion is supported by ultraviolet spectroscopic data, i.e., by comparison of the spectrum of the parent compound with those of its methyl derivative^.^? The infrared spectrum of
ari
OH
I
H [781
I
Me [801
H [ 791
I
Me [811
1-methyloxindole-3-aldehyde is in accord with its formulation as 80 and/or 81.85 I n 1883, the hydroxy structure 82 was assigned to indoxyl on the basis of chemical evidence.88 More recently, however, the infrared spectra of 1-acetyl- and 1-methyl-indoxyl measured in chloroform indicated that the 0x0 form 83 (R = Ac, Me) greatly predominates,4°pso "E. D. Bergmann, J. A m . Chem. SOC. 77, 1549 (1955). " P . L. Julian, J. Pikl, and F. E. Wantz, J. A m . Chem. SOC.57, 2026 (1935). = A . E. Kellie, D. G . O'Sullivan, and P. W. Sadler, J. Chem. SOC.p. 3809 (1956). Mme. Ramart-Lucas and Mlle. Biquard, Bull. S O C . chim. France 2, 1383 (1935). 88 A. von Baeyer, Ber. deut. chem. Ges. 16, 2188 (1883). BeH.C. F. Su and K . C. Tsou, J. A m . Chem. SOC.82, 1187 (1960).
TAUTOMERISM: 111. &MEMBERED RINGS-1
HETERO ATOM
19
but the presence of an acetyl group in the 2-position has been shown to cause enolization to The important dyestuff-intermediate, indigo white, is usually formulated as 85 (see, e.g., reference go), but
Ho@ J
I
H
R
~4 1
apart from “general phenolic character” there is little definite evidence to support this formulation.
I. OTHER COMPOUNDS WITH POTENTIAL HYDROXYL GROUPS Infrared and ultraviolet spectral comparisons by Grob and Hoferel show that the position of the equilibrium 86 e 87 predominantly favors 87.
OPT.S. Stevens, in “Chemistry of the Carbon Compounds” (E. H. Rodd, ed.), Vol. IVB, pp. 1081, 1094. Elsevier, Amsterdam, 1959. O’C. A. Grob and B. Hofer, Helv. Chim. Actu 36, 847 (1953).
20
A. R. KATRITZKY AND J. M. LAGOWSKI
111. Compounds with Potential Mercapto Groups Early work on thiophenethiols has been summarized by Hartough19? although few conclusions were reached concerning their tautomerism. Caesar and Brantong3 concluded from infrared spectral data that 3-mercaptothiophene existed a t least partly in the thiol form but that the 3,4-dithiol existed completely as the dithione 88. Gronowitz and his associates have recently made a definitive investigation of a series of thiophene-2- and -3-monothiols using nuclear magnetic
r e ~ o n a n c e ~and ‘ ~ ~infrared ~ ~ ~ ~ s p e ~ t r o s c o p y .Their ~ ~ ~ ~results ~ show convincingly that all of these compounds exist predominantly in the thiol form, both as the pure liquid and in cyclohexane solution. Mercapto-pyrr~les~’and -indo1esg8 have been characterized, but little is known about their tautomerism.
IV. Compounds with Potential Amino Groups The amino form is usually much more favored in the equilibrium between amino and imino forms than is the hydroxy form in the corresponding keto-enol equilibrium. Grob and Utzingerg9suggest that in the case of a-amino- and a-hydroxy-pyrroles, structure 89 increases the mesomeric stabilization and thus offsets the loss of pyrrole resonance energy, but the increase due to structure 90 is not sufficient to offset this loss. Similar reasoning may apply to furans and
’’H .
D. Hartough, “Thiophene and Its Derivatives,” pp. 42S429. Interscience, New York, 1952. “P. D. Caesar and P. D. Branton, Ind. Eng. Chem. 44, 122 (1952). M R .A. Hoffman and S. Gronowitz, Arkiv Kemi 16, 515 (1961). 95R.A; Hoffman and S. Gronowite, Arkiv Kemi 16, 563 (1961). MS. Gronowitz, P. Moses, and A. B. Hijrnfeldt, Arkiv Kemi 17, 237 (1961). ‘“P.Pratesi, Atti accad. Lincei 16, 443 (1932); Chem. Abstr. 27, 2442 (1933). ”W. C. Sumpter and F. M. Miller, “Heterocyclic Compounds with Indole and Carbarole Systems,” pp. 53-54. Interscience, New York, 1954. w C .A. Grob and H. Utzinger, Helv. Chim. Acta 37, 1256 (1954).
TAUTOMERISM: 111. &MEMBERED RINGS-1
HETERO ATOM
21
A
H
~ 9 1
[go1
thiophenes. The limited experimental evidence available certainly demonstrates that amino compounds are more stable as such than their hydroxy analogs.
A. AMINOFURANS The substituted 3-aminofurans (91, R = H, Me) resinify in air, can be diazotized (structure 91), and are easily hydrolyzed (structure 92 ? ) . l o oThe infrared spectra of both 2- and 3-acetamidofuran show a strong NH stretching band indicating that these compounds do, indeed, exist in the acetamido form.lol C0,Et m 2
Me [gza]
The a-aminobenzofuran 92a exists in the amino form shown, as evidenced by infraredlOlaand proton resonance spectra.lOlb ImH. B. Stevenson and J. R. Johnson, J. Am. Chem. SOC. 59, 2525 (1937). '"R. Kuhn and G. Kriiger, Chem. Ber. 89, 1473 (1956). lolaJ. Derkosch and I. Specht, Monatsh. Chem. 92, 542 (1961). lolbA. R. Katritzky and J. Derkosch, Monntsh. Chem. 93, 541 (1962).
22
A. R. KATRITZKY AND J. M. LAGOWSKI
B. AMINOTHIOPHENES The tautomerism of 2- and 3-aminothiophenes was mentioned by Hartough in his review of thiophenes,lo2but the first definite evidence became available in 1961 when Hoffman and Gronowitz’* showed conclusively by nuclear magnetic resonance spectroscopy that these compounds both exist in the amino form. In agreement with this finding, 3-aminothiophene generally behaves as an aromatic amine.lo3
C. AMINOPYRROLES Aminopyrroles are usually formulated as such because they are quite strong bases and do not easily lose ammonia by h y d r o l y s i ~ . ~ ~ ~ , ~ ~ ~ I n agreement with the amino formulation, the infrared spectrum of the a-aminopyrrole 93 (R = H) contains a vNH, doublet and a band near 1660 cm-l corresponding to the NH, deformation frequency, and the infrared spectrum of the acetamino derivative is in agreement with the structure 93 (R = Ac) .88 However, a stable imino compound, probably with structure 94, has been isolated.56
I
CH,CH (OEt), ~ 3 1
I
\
H
CO,But
[941
The relative stability of 2,5-diaminopyrrole- (95) and 2,5-diiminopyrrolidine-type (96) structures for succimidines has not yet been c1arified,100-108 although ultraviolet .spectral data indicate that forms of type 97 are probably unimportant.loD BanfieldllO has discussed H. D. Hartough, “Thiophene and Its Derivatives,” pp. 228-240. Interscience, New York, 1952. Io3E.Campaigne and P. A. Monroe, J. A m . Chem. Soc. 76, 2447 (1954). l‘C. A. Grob and P. Ankli, Helv. Chim. Acta 33, 273 (1950). losH. Fischer and H. Orth, “Die Chemie des Pyrrols,” Vol. 1, p. 110 ff. Akademische Verlag., Leipzig, 1934. lWG.E. Ficken and R. P. Linstead, J. Chem. Soc. p. 3525 (1955). ’O‘J. Schurz, A. Ullrich, and H. Bayzer, Monatsh. Chem. 90, 29 (1959). l m J . A. Elvidge, Chem. Sac. (London), Spec. Publ. N o . 4, p. 28 (1956). lO9 J. A. Elvidge and R. P. Linstead, J. Chem. Sac. p. 442 (1951). llOJ. E. Banfield, J. Chem. SOC.p. 2098 (1961).
TAUTOMERISM: 111. &MEMBERED RINGS-1 HETERO ATOM
H,N
rn N I H
HN QNH I
NH,
HN
H
23
NHZ
HZN. H,N
"H
/C=N [991
the fine structure of guanidino derivatives of type 99.
D. AMINOINDOLES Z-Aminoindole was initially assigned structure 100 (R = H) on chemical grounds"' and later structure 101 was erroneously assigned because the ultraviolet spectrum of its hydrochloride was similar to that of oxindole.llz In 1956, Kebrle and Hoffmannlls established the
Me
"'R. Pschorr and G. Hoppe, Ber. deu.t. chem. Ges. 43, 2543 (1910). "'H. Rinderknecht, H. Koechlin, and C . NiemanQ, J. Org. Chem. 18, 971 (1953). J. Kebrle and K. Hoffrnann, Helv. Chim. Acta 39, 116 (1956).
24
A. R. KATRITZKY AND J . M. LAGOWSKI
predominance of structure 102. 2-Amino-l-methylindole (100, R = Me) exists largely in the amino form in ethanol (see later), since its ultraviolet spectrum differs from that of 103 (R = Me). The ultraviolet spectrum of 2-aminoindole differs from both those of 100 (R = Me) and of 103 (R = Me) ; therefore, i t probably exists as 102. All three bases show similar ultraviolet spectra in acidic solutions indicating that they form similar cations, i.e., of type 104, and therefore the pK, method could be used to study these equilibria. Under all conditions, form 102 was found to be the most stable tautomer by a factor >lo. The equilibrium between 103 (R = H) and 100 (R = Me) is displaced toward 103 (R = H), the displacement increasing with increasing polarity of the solvent; a comparison of the ultraviolet spectra led to the same conclusion. Infrared spectra support these structural assignments. The acylamino compounds 105 ( R = H, Me, Ac)l13 and the isoindole derivative 106lI4 also exist in the forms shown on the basis of infrared and ultraviolet spectra1 evidence.
E. POTENTIAL AMINODERIVATIVES OF ISOINDOLE Ultraviolet spectral comparisons indicate that structure 107 predominates over 108 when R = H or OH, but that 107 is the predominant form when R = ~ r y 1 . lSimilarly, ~~ 109 predominates over 110 by a large factor when R = H , OH, or Me, and by a smaller factor when R is a higher alkyl group, but 110 predominates when R
is an aryl group. (For a discussion of guanidino derivatives corresponding to 110, see reference 110.)
V. Compounds with Potential Methyl Groups 2-Methylene-2,5-dihydrofuran (111) , which has been isolated by Rice,116is converted by acid into 2-methylfuran.
I w M . E . Baguley and J. A. Elvidge, J . Chem. Soc. p. 709 (1957). lX5P.F. Clark, J. A. Elvidge, and J. H. Golden, J . Chem. SOC.p. 4135 (1956). IlaH. L. Rice, J . Am. Chem. Soc. 74, 3193 (1952).
TAUTOMERISM : 111. &MEMBERED RINGS-1
HETERO ATOM
8
0
-
25
NHR
c
RN
VI. Other Substituted Pyrroles A. VINYLPYRROLES The tautomeric equilibrium between 112 and 113 has been studied by Treibs and his associate^.^^^^'^^ Most pyrromethenes take form
113 unless both rings carry electron-withdrawing substituents in which case structure 112 is predominant. Common cations of type 114 are favored by both tautomers.
Treibs and F. Reitsam, Ann. Chem. Liebigs 611, 194 (1958). ,"A. Treibs, E. Herrniann, E. Meissner, and A. Kuhn, Ann. Chem. Liebigs 602, InA.
153 (1957).
26
A. R. KATRITZKY AND J . M. LAGOWSKI
B. NITROSOPYRROLES Chemical evidence has been advanced for the formulation of p nitrosoindoles as either 115 or 116 (see reference 119 and references therein), but ultraviolet spectral comparison with both methylated forms clearly indicated that 116 was favored. Infrared spectral data are also considered to support structure 116.'*O
H [115J ~~
[lltij
~
I*'N. Campbell and R. C. Cooper, J. Chem. SOC.p. 1208 (1935). I2'F. Piozzi and M. Dubini, Gazz. chirn. ital. 89, 638 (1959).