The chemistry of enaminones, diazocarbonyls and small rings: our contribution

REVIEW

The chemistry of enaminones, diazocarbonyls and small rings: our contribution

Concetta Montanile Kascheres

Instituto de Química, Universidade Estadual de Campinas, CP 6154, 13083-970 Campinas-SP, Brazil

Universidade Paulista, Instituto de Ciências da Saúde, Avenida Independência, 412, 130087-050 Sorocaba, SP, Brazil

ABSTRACT

After thirty two years in Brazil and retired from Universidade Estadual de Campinas the author wishes to present this account, a summary of a large part of the research in synthetic methodology developed by the research groups of Albert J. Kascheres and the author at Universidade Estadual de Campinas Chemistry Institute. Contributions have been made to the area of enaminones, diazocarbonyls, cyclopropenones and azirines.

Keywords: enaminones, diazocarbonyls, cyclopropenones, azirines

RESUMO

Depois de trinta e dois anos no Brasil e de aposentada da Universidade Estadual de Campinas, a autora gostaria de apresentar este resumo representando uma grande parte da pesquisa em metodologia sintética desenvolvida pelos grupos de pesquisa de Albert J. Kascheres e da autora no Instituto de Química da Universidade Estadual de Campinas. Contribuições foram feitas nas áreas de enaminonas, diazocarbonilas, ciclopropenonas e azirinas.

1. Enaminones

Enaminones are chemical compounds consisting of an amino group linked through a C=C to a carbonyl group. They are versatile synthetic intermediates that combine the ambident nucleophilicity of enamines with the ambident eletrophilicity of enones. They are typical push-pull ethylenes in which the amine group pushes and the carbonyl pulls electron density. The carbonyl group, conjugated to the enamine moiety, gives this system enough stability to be easily prepared, isolated and stored under atmospheric conditions at room temperature. The chemistry of the enamino carbonyl group (A) is potentially an area of considerable scope when one considers that there are present in this moiety three nucleophilic sites (a, c and e) and two electrophilic sites (b and d).^1,2

There are structural questions that have to be addressed concerning these systems; the first involves tautomerism. Primary and secondary acyclic enaminones can exist in three tautomeric forms (I-III), as shown in Figure 1. There has been collected a great deal of evidence using NMR, UV, IR and X-ray analyses, showing that the ground state of enaminones is best characterized by tautomeric form I.³ There is also evidence that cyclic enaminones exist primarily in the enaminoketo form (I).⁴ Our theoretical calculations comparing the AM1 and MNDO methods showed that AM1 gives results consistent with this experimental data.⁵

Once the correct tautomer has been established, the geometric form of enaminones must be determined. These compounds are known to exist in the following geometric forms: Z,s-Z; Z,s-E; E,s-Z; and E,s-E (Figure 2). The conjugation of the system facilitates the interconversion between the Z and the E forms. In general, spectroscopic methods have shown that primary and secondary enaminones exist predominantly, and in many cases completely, in the Z,s-Z form in solution in apolar solvents. This allows for intramolecular hydrogen bonding. An increase in solvent polarity, particularly solvents that form strong intermolecular hydrogen bonds, increases contribution of the E forms. Tertiary enaminones tend to adopt the E forms which are sterically less hindered.⁶

1.2 Reaction with electrophiles

For most of our studies the following enaminones, presented in Figure 3, were used:

1.2.1 Reaction with diphenylcyclopropenone

Cyclopropenones are strained ring ambident electrophiles with a tendency to form ring opened products. In reaction with nucleophiles one has to consider the possibility of carbonyl or conjugate addition.

The reaction of diphenylcyclopropenone (8a) with a variety of tertiary enamines (9) was suggested to proceed through ylide (10) and ammonium enolate (11) intermediates to afford the C-N insertion products (12) (Scheme 1).⁷ Intermediates analogous to 10 and 11 were also proposed in the reaction of tertiary enaminones (9, R²=COR).⁸

Although primary and secondary enaminones were known at the time of the above studies, their utilization as mechanistic probes in the above reaction, where rapid proton transfer in intermediates might permit the isolation of interesting heterocycles, was not explored. Therefore we undertook the study of primary and secondary enaminones with diphenylcyclopropenone (8a).

The reaction of the acyclic enaminones 1a, b, 2a and 3 with diphenylcyclopropenone (8a) led to the formation of 5-functionalized 1,5-dihydropyrrol-2-ones (13) for enaminones 1 and 2. The results permitted an alternative representation (Scheme 2) of the reaction of 8a with acyclic enaminones as a cycloaddition at NC_b of the enaminone to afford intermediate (14), which could also account for the products 15 obtained from tertiary enaminones.⁹ This would require that the reaction be initiated by attack at the carbonyl carbon of 8a.

According to perturbation theory, the major contributions to bond making when two reagents approach each other involve opposite charge attraction (i.e., electrostatic interactions) at the reaction sites and interactions between the filled orbitals of one reagent and the unfilled orbitals of the other, with HOMO's and LUMO's corresponding to the most important interactions because of the proximity of their energy levels. When the HOMO/LUMO interaction is the major factor governing reactivity, the reaction is said to be frontier-orbital controlled.¹⁰ Since 8a acts as an electrophile, its LUMO should be important if the reaction is frontier-orbital controlled. The LUMO coefficients for this system were obtained using the AM1 approach. The results of this calculation, 0.455 for phenyl-C versus 0.0001 for C_CO¹¹ suggest that the reaction should involve the phenyl-C of 8a. Thus, another suggestion (Scheme 3) is now presented for this reaction.

The cyclic enaminones 4 proved to be much less reactive toward 8a. The principal product of the reaction of 4a, a 1:1 adduct, was assigned the 1,5-dihydro-4H-pyrrol-4-one structure 16. In the case of 4b, the 2:1 product (17b) was the principal cycloadduct. A 1:1 adduct was also isolated, and was eventually identified as the 1,5-dihydro-2H-pyrrol-2-one 18. Thus, of the one to one adducts isolated in this study, 16 was the only 1,5-dihydro-4H-pyrrol-4-one observed.⁹ While products 13 seem to envolve reaction of the enaminone at the C_a position, these products suggest reaction of the cyclic enaminone 4 at N (Scheme 4).

Compound 13 or a derivative of it was envisioned as a starting point for preparing bicyclic [3.3.0] systems. Thus, the presence of an activated methylene in the C-5 substituent might permit construction of a second ring through intramolecular aldol condensation involving an appropriately situated carbonyl group on nitrogen. Therefore, enaminone 19 was prepared quantitatively by the reaction of the aminoacetal 20 with acetylacetone. The reaction of 19 with diphenylcyclopropenone (8a) afforded derivative 21 in 71% yield. Hydrolysis of 21 with concentrated hydrochloric acid afforded compounds 22 and 23 (Scheme 5). Compounds 22 and 23 contain the nucleus present in the pyrrolizidine alkaloids. Many pyrrolizidine derivatives are known,¹³ but few with a carbonyl group in the 2 position of the rings as in 22 and 23 have been documented.¹⁴

Another approach was to convert the ketone fragment in 13 to other versatile functional groups, such as a primary amine. A well established procedure toward this end involves submitting the oxime to Raney nickel generated in situ.¹⁵ However, the reduction of oximes 24a-f under these conditions gave only bicyclic compounds 26a-f and 27a-f, obtained as single diastereomers (Table 1), while hydrogenation using W7 Raney nickel produced only amines 25 (as 1:1 mixtures of diastereomers) which upon acetylation formed 28d-f. The high diastereoselectivity observed in the transformation of 24 to 26 was understood in terms of participation of intermediate 29 as shown in Scheme 6. Formation of the new C4-N8 bond requires approach of N8 from below the heterocycle plane thus favoring transfer of hydrogen from the catalyst surface to the imine fragment via the Re face in the stereoisomer shown. Competing reductions to 30 and, eventually to 25, without selectivity, occurs when R¹ in 24 is small.¹⁶ Compounds 26 and 27 are pyrrrolo[3,2b]pyrrole derivatives. Several examples containing this nucleus have been cited,¹⁷ but few oxo derivatives have been documented.¹⁸

The principal difference in reactivity of enaminothiones 31 when compared to enaminones resides in the observed heterodiene character of the enaminothione in Diels-Alder reactions with electron deficient olefins, as well as behavior as a nucleophile almost exclusively through sulfur.¹⁹

The observed participation of 8a as a dienophile in reaction with (diethylamino)butadiene (equation1)²⁰ together with our results with enaminones (Scheme 3 and 4) prompted us to examine the reactivity of enaminothiones 31, wherein the effect of the b-thione substituent might be compared with that of the b-ketone and b-vinyl derivatives previously studied.

The reactions of 8a with 31a-c occurred slowly in refluxing benzene to afford 4-methylenecyclopentenones 32a-c in good yield (Scheme 7). The most straightforward mechanistic interpretation of the formation of 32 involves initial nucleophilic attack of 31 at the carbonyl carbon (C₁) of 8a through C_a, with subsequent ring expansion to the thione carbon. The problem here is that enaminothiones are not known to react as nucleophiles through C_a and, as discussed above, C1 of 8a should not be the reactive electrophilic center. With the objective of gaining further insight into the electronic distribution in 31, an AM1 calculation as implemented in the AMPAC package was performed on geometrically optimized 31a in the Z, s-cis-configuration. For the purpose of comparison, a similar calculation was performed on the oxygen analog. The fronteir orbital treatment calls attention to the importance of the HOMO of 31a in the reactions of this system as a nucleophile. This HOMO is essentially a nonbonding orbital on sulfur. In contrast, the corresponding orbital of the oxygen analog is the p-system with largest coefficients at C_a and N. Interestingly, the chemical behavior of these systems toward electrophiles, reflects exactly these differences. Thus, as mentioned above, enaminothiones react through sulfur, while enaminones react through C_a or N. A mechanism for the formation of 32 wherein the nucleophilic site is sulfur, is presented in Scheme 7. While more round-about, this pathway is more coherent with regard to the expected behavior of 31.²¹

1.2.2 Reaction with diazocarbonyls

Diazocarbonyls, like enaminones, have added stability when compared to the parent compounds because of the carbonyl group. They are known to react through the starting compound, through ketocarbenes formed by the loss of molecular nitrogen and through ketenes formed by loss of nitrogen followed by the Wolff rearrangement. Ketocarbenes and ketenes are electrophilic intermediates.

As already mentioned, enaminones have three nucleophilic and two electrophilic centers. A systematic study of diazocarbonyl compounds with enaminones was undertaken based upon the variety of reactions that would be possible which could lead to new and interesting compounds.

Benzyldiazoketone (33a) was reacted with enaminones 1 and 2 under thermal conditions. The reactions occurred through the ketene intermediate to form a-acylenaminoketones (34) as the principal product; i.e., a product of reaction at C_a of the enaminone. Amides 35 and in the case of 2a, compound 36, were also formed (Scheme 8). Compounds 35 and 36 suggest reaction at the nitrogen of the enaminone.²²

In all cases, competitive formation of decomposition products of 33a occurs. It was observed that for both enaminoketones 1a-c and enaminoesters 2a-c, the total yield of products and the relative yields of the products of reaction at the C_a carbon [proportion 34:(35 + 36] increase in going from R³ = H to Me to t-Bu while the yield of decompostion products of diazoketone 33a decreases, suggesting that the order of reactivity for the enaminones is 1a > 1b > 1c and 2a > 2b >2c. For a given substitution on nitrogen, the total yield of products is greater for the enaminoesters. Molecular orbital calculations using the HAM/3²³ semi-empirical method shows that for equal substitution on nitrogen, the HOMO energy of the enaminoester is greater than for the enaminoketone and within each series the HOMO energy follows the order R³=H < R³=Me < R³=t-Bu. If one considers that by frontier orbital theory a higher energy HOMO implies greater reactivity of the enaminone as nucleophile, we can see an excellent agreement between HAM/3 HOMO energies and the reactivity observed.²²

The HOMO of the enaminones corresponds to a p molecular orbital with the largest coefficients on the C_a and nitrogen (see Scheme 9). The absolute magnitude of the two coefficients is somewhat greater at C_a. Therefore, the products formed are consistent with a HOMO-controlled process in which initial attack by diphenylketene occurs at these centers (Scheme 9).

The primary and secondary acyclic enaminones 1 and 2 are known to exist in the Z,s-Z form. The reactions of diphenylketene with the cyclic enaminones 4, 5 and 6 were also carried out to see the effect of fixing the enaminone in the E,s-E form. The principal products obtained in the case of the cyclic enaminones 4, 5 and 6 correspond to reaction at nitrogen; i.e. products 37. In the case of 4a there was a small amount of product of reaction at C_a (38, Table 2).

For these fixed E,s-E enaminones, intramolecular hydrogen bonding is not possible, which can decrease conjugation when compared to acyclic Z,s-Z enaminones. This would increase the electron density on nitrogen and thus favor reaction at this site. At the same time, attack at the C_a carbon might be slowed down because of the development of substantial 1,3-diaxial interaction in the transition state, which we expect to be important for the volumous diphenylketene. Further information was obtained using AM1 geometric optimization, which shows that while the nitrogens of the Z,s-Z acyclic enaminones are planar, those of the cyclic systems are not. This is an indication of the decreased conjugation in the cyclic enaminones. HAM/3 molecular orbital calculations show a high energy second HOMO which corresponds to a non-bonded electron pair on oxygen for the cyclic enaminones. Thus, this orbital would also have to be considered in any analysis of reactivity. In fact, this might explain the greater tendency for cyclic E,s-E enamino ketones to give high yields of O-alkylated salts upon alkylation and to give O-acylated products or Ca-acyl derivatives via unstable O-acylated intermediate salts upon acylation. Calculation on compound 1a using standard geometric models and fixed in the four possible geometric forms show that the high energy of the second HOMO is a function of the s-E conformation (Table 3). Thus, to explain the reaction of the cyclic enaminones 4, the possibility of prior reaction of the diphenylketene on oxygen to form an unstable intermediate that then reacts further with another molecule of the enaminone cannot be dismissed. However at some point during the reaction a greater preference for reaction on nitrogen would have to be explained.²²

Under Cu(acac)₂ catalytic conditions the acyclic enaminones 1and 2 react with 33a to form pyrroles 40 by the inital formation of the ketocarbene of 33a (i.e. 39a). Upon further investigation with the ketocarbenes from 33b and 33c, we discovered that these pyrroles were also formed by initial reaction at the a-carbon of the enaminone as can be seen from the formation of 41 as the main product from 39b and 42 as the only product from 39c (Scheme 10).

In the case of cyclic enaminones 4a, 5 and 6, compounds 43 were initially formed and then converted to pyrroles 44 (except for 43 formed from 6) by the action of base (Scheme 11). Again the acyclic enaminones react mainly through C_a and the cyclic ones through nitrogen (see Schemes 3 and 4).²⁴

To the best of our knowledge, these results represented the first example of the use of ketocarbenes in pyrrole synthesis. Therefore with this in mind we decided to look at other diazo systems to test the generality of the reaction and to be able to synthesize more complex pyrroles. The reaction of 2-diazo-1-indanone (46) was examined. This compound is prepared from 2-oxime-1-indanone (45) which forms 3-carboethoxy-2-methyl-1,8-dihydroindene[2,1-b]pyrrole (47) with ethyl acetoacetate under Knorr conditions, followed by methylation. The reaction of 46 could form the same product 47 with enaminoester 2b if the reaction occurred on the nitrogen of the enaminone or it could form the isomer 1,4-dihydroindene[1,2-b]pyrrole 48, if it reacted on C_a. The product actually formed was 48 (Scheme 12) which again shows reaction of the acyclic enaminone at C_a and the possibility of forming isomeric systems from the same starting material; i.e. the oxime 45 (used to prepare 46).²⁵

We next looked at the reactions of ethyl diazoacetate 49 under catalytic conditons. Again pyrroles 51 were formed from the carboethoxycarbene 50, but they were unexpected ones and the mechanism of their formation was not obvious. Results from varying the structures of the enaminones allowed us to determine that in these pyrroles the C(3)-CO₂Et structural fragment comes from the carboethoxycarbene formed from 49 and that the fragment originating from the enaminone is C(4)H-C(5)-R¹NR². The position of the C3-CO₂Et fragment in the pyrrole indicates that the initial step of the reaction mechanism occurs through an electrophilic attack of the carbene at the C_a of the enaminones, leading to the net result of C_a-H insertion. To investigate the origin of the C(2)-Me fragment, the reaction of 49 with enaminoketone 1d (no substituent at both C_a and Cb) was performed. Work-up of the reaction mixture revealed pyrrole 51c as the only product which could be isolated and identified indicating that the C(2)-Me fragment originated from a second molecule of carboethoxycarbene. These results show that the COR groups from 1 and 2 are eliminated during the course of the reaction (Scheme 13). Even with this data in hand, a reasonable mechanism for the process could not be proposed. However, the driving force for the reaction is undoubtedly the formation of the stable aromatic pyrrole system.²⁶

1.2.3 Reaction of 3-diazo-1,3-dihydro-2H-indol-2-ones and related compounds with enaminones

The reaction of 3-diazo-1,3-dihydro-2H-indol-2-one (52a, Table 4) with 1b led to the easy isolation of a product that precipitated from the reaction mixture and which we identified as a pyrazoloquinazolinone (60% yield) (Scheme 14).

We had a characterization problem at hand though, i.e. which isomer was formed (53a or 53a').

Mass analyzed ion kinetic energy spectroscopy (MIKES), a mass spectrometric technique which determines the products of metastable ions formed in the second field free region of a mass spectrometer with reverse Nier-Johnson geometry,²⁷ was used in a first attempt to establish the structure. The MIKES spectrum (Figure 4) showed the loss of CH₃COCN, which would only be reasonable for isomer 53a.²⁸ However, the intensity of this transition was very low and this was a relatively new type of spectroscopy. We had doubts as to the reliability of the results. Thus we waited until a detailed nmr study could be accomplished and in fact COLOC confirmed 53a as the correct isomer. A more detailed look at the reaction mixture produced the triazole (54) as another product in low yield (6%).²⁹ Both systems are of interest from the biological viewpoint. Thus, pyrazolquinazolinones belong to a class of compounds with antiallergenic and anti-inflammatory activity³⁰ and there are a number of examples of 1,2,3-triazoles with important biological activity such as the inhibition of the growth, mobility, and adhesion of cancerous cells in cultures and in rats³¹, herbicides,³² fungicides,³³ antiallergenics,³⁴ anticonvulsants⁵² and interference in the metabolism of prostaglandins.³⁶

This reaction shows the diazocarbonyl compound reacting by way of the starting material. A detailed study with substituted 3-diazo-1,3-dihydro-2H-indol-2-ones 52b-f showed that electron withdrawing groups on the diazo compound favored reaction (Table 4). Both compounds could be explained by frontier orbital theory considering an initial reaction of the C_a of the enaminone (HOMO) on the terminal nitrogen of the diazo compound as shown in Scheme 14 (LUMO of the diazo compounds show large coefficients on the two diazo nitrogens). The reaction was more easily directed toward the formation of the triazoles.²⁹ The most important methods of 1,2,3-triazole synthesis involves reaction of azides with acetylenes. These triazoles would require the use of N-alkyl azides which are hazardous to work with.

We envisioned the use of this novel reaction in the preparation of bicyclic 1,2,3-triazole systems. The ability to form more highly nitrogenated analogs of aromatic compounds, like pyrroles and imidazoles, becomes important when one considers that as the number of heteroatoms in the ring increases, the tendency toward electrophilic attack decreases as does the reactivity of the benzylic leaving groups toward substitution. Pyrrole metabolites of pyrrolizidine alkaloids represent examples of compounds in which the high reactivity (and toxicity) of the pyrrole ring precludes their use in cancer chemotherapy. Thus, various attempts have been used to synthesize triazole analogs which modulate the reactivity downward.³⁶ With this in mind the reactions of enamino esters 56 (Scheme 15) and cyclic enamino ketones 4b and 5 were undertaken.³⁷

The reaction of 52f with 56a (n=1) formed 3-carboethoxy-5,6-dihydro-4H-pyrrolo-[1,2,3]-triazole 57a in 78% yield. This product can be used as a synthon to prepare triazole analogs of the tumor inhibitory pyrrolizidine alkaloid pyrrole metabolites.³⁸ Enamino ester 56b (n=2) reacts similarly to form 3-carboethoxy-4,5,6,7-tetrahydro-[1,2,3]- triazolo-[1,5a]-pyridine (57b) in 83% yield. This is a triazole analog of the indolizidine alkaloid skeleton. Enamino ester 56c (n=3) reacts with 52f to form 3-carboethoxy-5,6,7,8-tetrahydro-[1,2,3]- triazolo-[1,5a]-azepine (57c) in 73% yield. In the reactions of 4b and 5, there is the formation of 58 in approximately 50% yield. Dihydroisatin 55c was isolated in many of the reactions.³⁷ This compound can be oxidized to the starting dinitroisatin which can then be used to form 52f, the starting diazocompound. Thus, even though most of the atoms of 52f are not incorporated in the triazoles, a recyclable product is obtained.

1.2.4 Reaction of enaminones with diazoquinones. Non-linear optical properties

Quinone diazides are compounds that have intermediate reactivity between diazonium salts and aliphatic diazoketones.³⁸ They can react by either loss or retention of molecular nitrogen, depending on reaction conditions. An example of the first case involves the formation of dihydrobenzofurans by a regiospecific route in reactions of o-quinone diazides with vinyl ethers under thermal conditions.³⁹ On the other hand, retention of nitrogen in azo-coupling reactions is often observed.⁴⁰ This type of reaction is widely used in industry for the synthesis of hydroxyazo dyes.⁴¹

Since quinone diazides, especially nitro-substituted ones, are very sensitive to light we decided to try a one-pot procedure in the reactions with enaminones in order to avoid contact of the reacting mixture with light. The quinine diazides 60 were generated in the reaction media by nitrosation of the proper o-aminophenol to form diazonium salt 59 which is then neutralized to form 60. The quinone diazides formed were reacted with enaminones 1 and 2 to produce the novel azo-enaminones 61 which were not formed when 59 was reacted directly with the enaminones. In that case deaminated azo compounds 62 were formed instead (Scheme 16).⁴²

AM1 geometry optimization of 61 indicate a tendency for two intramolecular hydrogen bonds with the azo group and a planar structure for the azo-conjugated skeleton. X-ray crystallographic data for 61d⁴³ indicate good agreement between X-ray and AM1 calculated geometries. Based on these data all azo-compounds synthesized are assumed to have the two hydrogen bonds linked to the azo group. The planarity of the azo skeleton leads to extensive conjugation throughout the molecule, and when the phenyl ring is attached to electron-withdrawing groups, like chlorine or nitro, a push-pull conjugation is observed.

Organic compounds containing push-pull conjugation can present large optical second-order molecular polarizabilities and have important applications in second-order nonlinear optical (NLO) devices. Delocalization and intramolecular charge transfer effect of p-electrons in these molecules can cause large and fast nonlinear responses.⁴⁴ Extensive calculations using semi-empirical models have been reported for many varieties of molecular architectures, and many of these calculations have not only determined numerical values for nonlinear response, but have also dealt with interpretation of the molecular geometry and the electronic structure of such organic conjugated molecules.^45,46

We decided to evaluate the molecular polarizability of these azo compounds theoretically because we did not have conditions for measuring them experimentally. The AM1/finite-field method was chosen as our first approach since it is reasonably fast and has been used successfully for determining trends of molecular hyperpolarizabilities.⁴⁷ Results from these calculations show that the push-pull conjugation is essential for a significant value of the molecular coefficient of hyperpolarizability, b. Thus, compound 61a has no electron-withdrawing groups on the phenyl and has the smallest value for the molecular coefficient of hyperpolarizability, b. Intermediate values were found for 61b,c, and e, and for compounds 62b,c which have a poor donating group (OH). The other six azo-enaminones seem promising for second harmonic generation; they have significant values for second harmonic generation and are almost planar. The p-nitroazoenaminones have the largest value for b and therefore show most promise in non-linear optics. In conclusion, our procedure allowed us to obtain novel azo compounds in good yields, especially azo-enaminones containing push-pull conjugation. In addition, we proposed the potential usefulness of the push-pull azo-enaminones (61) in nonlinear optics, as second harmonic generators, based on theoretical finite-field static calculations.⁴²

1.3 Reaction with nucleophiles

Enaminones react with hydrazines to form substituted pyrazoles.⁴⁸ As already seen, diazoketone 33a reacts with enaminones via ketenes under noncatalytic thermal conditions to form nucleophilic addition products. The same reaction occurs with 33b and 33c to form the a-acylenaminoketones 34b [3-acetyl-1-phenyl-1-methyl-4-(methylamino)-3-penten-2-one] and 34c [3-acetyl-1,1-dimethyl-4-(methylamino)-3-penten-2-one]. The study of the reactivity of a-acylenaminoketones 34a-c interested us because of the differences in the two ketonic carbonyls, especially since X-ray analysis of 34a showed that 1,1-diphenylacetyl group is perpendicular to the planar methylaminopentenone system.⁴⁹ Both ab initio 6-31G** and AM-1 geometry optimizations of 34a-c give results which go along with the X-ray analysis of 34a. ⁴⁹ Solvent effects were also of interest to us. Thus, benzene, methylene chloride, tetrahydrofuran, methanol and N,N-dimethylformamide were utilized in the reactions of a-acylenaminoketones 34a-c (Scheme 17) with hydrazine reagents (methylhydrazine, phenylhydrazine, p-nitrophenylhydrazine and hydrazine hydrate). Besides verifying which of the two carbonyls would preferentially be attacked during a nucleophilic attack, we wished to obtain information on the solvent dependence of the regiochemistry of the pyrazole formed. The reaction mixtures were submitted to gas chromatography/mass spectrometry analyses, in an attempt to identify all the products and possible intermediates formed during the reactions.

The formation of the principal pyrazoles 63 (the product formed by nucleophilic attack on the acetyl group), 64 (deacetylated pyrazoles, with the substituent R in the 5 position) and 65 (4-acetylpyrazoles formed by nucleophilic attack on RCO) can be explained by an initial Michael-type reaction (Scheme 17). The pyrazoles 64 and the deacetylated enaminones 66 (4-methylamine-2-pentenones) were formed by a deacetylation process. Small amounts of isomeric pyrazoles (67, isomer of 64 and 68, isomer of 65)) were also formed (perhaps by initial reaction on the carbonyl) together with acetamides (68) and 3,5-dimethylpyrazoles (70).

With methylhydrazine, the main product is 63, except for 34b and 34c in dimethylformamide, in which case, 65 is the main product. With phenylhydrazine, the formation of 64 becomes important, as is the case with p-nitrophenylhydrazine. With hydrazine hydrate, the main product is 65, except for 34a and 34b in methanol, in which case, 63 is the main product.⁵⁰

Simple eye inspection of the distribution of the eight products, shown above, especially those obtained in low yields, did not enable detection of consistent differences between solvents or nucleophiles. Therefore, in an attempt to understand how these factors could be correlated, we utilized principal component analysis (PCA).

Three principal components describe 87% of the total variance in the original data set. Thus, the pyrazole 64, which was obtained by a mechanism involving a deacetylation process was separated from pyrazoles 63 and 65 in the first principal component. The second component, on the other hand, has information about the structural characteristic of the a-acylenaminoketones. In this component 63 is separated from 65 which correspond to pyrazoles formed by attack at different carbonyls.

These reactions were separated into four groups, according to the nucleophile used in the reaction. The data suggest that the utilization of methanol, the only protic solvent used, favors nucleophilic attack on the acetyl carbonyl group to form 63. Products obtained via a deacetylation process were favored by a decrease in the nucleophilicity of the secondary amino group of the hydrazine.⁵¹

1.4 Small rings

We have already mentioned the reactions of diphenylcyclopropenone with enaminones and have seen that both these systems are ambident substrates. 1-azirines represent another versatile small ring system containing a nucleophilic reaction site on nitrogen and an electrophilic one on C2.

Thus, the reaction of diphenylcyclopropenone 8a with mono and disubstituted azirines 71 in refluxing toluene led to the formation of 4(1H) pyridones 72. These compounds can be pictured as forming by way of nucleophilic attack of the weakly basic azirine nitrogen on the C2 position of the electrophilic diphenylcyclopropenone ring followed by an electrocyclization, as illustrated in Scheme 18.⁵²

1.5 Cyclopropenone and derivatives

1.5.1 Reaction of cyclopropenones with aminoaromatics

A carbonyl addition product 73 (40% yield) had been identified in the reaction of diphenylcyclopropenone (8a) with pyridine⁵³ (equation 2). Interest in the chemical behavior of 8a made it desirable to explore the possibility of participation of a "conjugate addition" mode in the reacton of pyridine with appropriately substituting the nucleus with a second functional group capable of intercepting a reactive (ketene) intermediate. The 2-aminopyridine system (74) was chosen as a probe for this pathway.

Diphenylcyclopropenone underwent a smooth reaction with a variety of 2-aminopyridines at room temperature to form 75 and 76. Isomerization of compound 76 was observed to be solvent dependent. The lack of reactivity of aniline and other studies to understand the mechanism of formation of 75, led to a visualization of the reaction as a conjugate addition pathway involving initial nucleophilic attack of the ring nitrogen of 74 on the electrophilic cyclopropenone ring to eventually form 76 and 77 as shown in Scheme 19. The formation of 76 and 77 in this study apparently represented the first report of a 3,4-disubstituted 3,4-dihydro-2H-pyrido[1,2-a]pyrimidin-2-one.⁵⁴

Among the factors contributing to the facility of the ring-opening process to form 75, an unfavorable H-6:phenyl interaction and the tendency toward the benzenoid-like aromatic nucleus were mentioned.⁵⁴ Analogous reactions were later observed with amidines and guanidines.⁵⁵ The above results led to the conclusion that this reaction could be applied to the synthesis of 5,6-diphenyl-5,6-dihydrouracils by reaction of diphenylcyclopropenone with thiopseudoureas as masked urea fragments. Therefore, thiopseudoureas 78 were reacted with 8a and formed the expected dihydrouracils 79 as shown in Scheme 20. It is interesting to note that urea does not react with 8a under the reaction conditions.⁵⁶

With the objective of evaluating the potential of the above reaction in the synthesis of bicyclic systems, a study of the reactivity of 8a with 2-aminothiazoles (80) and related compounds was undertaken.⁵⁷ The reactions of 80 with diphenylcyclopropenone led to the formation of 81 in average to very high yields. In comparison with 76, 81a-c demonstrated a remarkable stability in solution. The formation of 81 may be visualized as occurring by way of a highly stereospecific and efficient intramolecular interception of a ketene intermediate 82, formed by initial conjugate addition as in the case shown above.⁵⁷

Unsymmetrical cyclopropenones such as methylphenyl-cyclopropenone 8b (Scheme 21) can react with nucleophiles at either of the two conjugate positions. Cyclopropenone 8b proved to be less reactive than 8a in reaction with 2-aminopyridines. Analogous pyridopyrimidinones 83 were formed and proved to be more stable than the diphenyl analog 76. These compounds (83) could be transformed to 84 in refluxing chloroform. An unstable primary product 85 was detected in the case of 74d and 74e. The reaction was first thought to occur by initial nucleophilic reaction of the ring nitrogen on the electrophilic C-Ph of the cyclopropenone as shown in Scheme 21.⁵⁸ However, HOMO-LUMO coefficients obtained from AM1 calculations of geometrically optimized 8b show that both HOMO and LUMO coefficients are largest at methyl-C (Table 5)⁵⁹ suggesting that reaction of nucleophiles with 8b is kinetically favored at methyl-C. Thus, the possibility that the reaction of 8b with 2-aminopyridines actually proceeds via an initial fast and reversible reaction at methyl-C in conjunction with a slow and irreversible ketene-forming sequence involving attack at phenyl-C, as shown in Scheme 22 has to be considered.⁵⁹

Reaction of 8b with pyrazole, wherein hydrogen transfer in intermediates involves no loss of heterocyclic aromaticity, shed new light on this question. An equimolar mixture of these reagents afforded ketone 86a (98% yield) together with unreacted 8b, confirming the participation of a cyclopropanone intermediate resulting from initial nucleophilic attack at methyl-C. A second equivalent of pyrazole may be incorporated via dipolar species 87. The viability of interception of 87 by other nitrogen nuclephiles was studied. The use of 2-aminothiazole, 2-amino-4-methylpyridine and o-diphenylenediamine (which does not react with 8b in the absence of pyrazole) formed 88, 89 (and 86a) and 90, respectively, as shown in Scheme 22.

With the objective of comparing these results with those involving the possible participation of a ketene intermediate, the above reactions of pyrazole were repeated with 8a in substitution of 8b. In the case of 2-aminothiazole, formation of previously reported 81 (78%) was observed, in addition to 91 (20%, mixture of diastereoisomers). In the presence of 2-amino-4-methylpyridine, the previously reported 75 (a ketene rearrangement product, 83%) was produced, in addition to 91 (11%) (Scheme 23). These results reflect the inherent nucleophilic superiority of aminothiazoles and aminopyridines, which may only be brought to light in irreversible ketene-forming processes. The extensive incorporation of pyrazoles in the competitive studies of 8b may be attributed to the formation of relatively stable cyclopropanone intermediates resulting from reaction at methyl-C.⁵⁹

1.5.2 Reaction of cyclopropenones with pyridinium N-Imines and other ylides

As a continuation of the study of cyclopropenones with pyridine derivatives, the reactions of cylopropenones with pyridinium N-imines and other ylides were examined. A ketene intermediate was invoked for the reaction of diphenylcyclopropenone (8a) with N -acylpyridinium imines (92a) which formed 2,4,5,-trisubstituted 6H-1,3-oxazin-6-ones (93).⁶⁰ The reaction of diphenylcyclopropenone (8a) with N-aminopyridinium iodide, in methanol, in the presence of a tertiary amine, led to the formation of 94 which can be visualized as occurring by way of a 1,2 or 1,4 addition of methanol to an iminoketene intermediate (95), formed by initial conjugate additiion of 92b on 8a with imine acting as a nucleophile, as shown in Scheme 24. The interception of 95 was also accomplished in aqueous dioxane to form deoxybenzoin, in formamide to form 96 (hydrolyzed to the known 97) and in aniline to form 98.⁶¹

When methylphenylcyclopropenone (8b) was reacted with pyridine N-imine 92b in methanol, formation of both enamino ester 100 and the 3H-pyrido[1,2b]pyridazine-3-one 101a was observed.

When the reaction was carried out in dichloromethane an intermediate 102 was isolated in addition to 101. On standing 102 was readily oxidized to 101. This reaction apparently involves participation of 92b not only as a nucleophile, as in the case of diphenylcyclopropenone, but also as a 1,3 dipolar compound. Thus, this work represents the first example of a possible 1,3-dipolar addition of a pyridine N-imine to a cyclopropenone, further illustrating the effects of substituents on the reactivity of the cyclopropenone ring.⁶² As a continuation to this study, methylphenylcyclopropenone was reacted with substituted pyridinium N-imino salts 99 in methylene chloride in the presence of triethylamine or potassium carbonate at room temperature. The results are summarized in Scheme 25. The reaction of di-n-propylcyclopropenone needed reflux conditions. The results show nucleophilic reaction for the diphenylcyclopropenone (Scheme 24) and cycloadditions for the methylphenyl and dipropylcyclopropenones (Scheme 26).⁶³

1.5.3 Reaction of cyclopropenones with N-acylamidine derivatives

N-acylamidines can be envisioned as enaminone aza analogs (an aza-enaminone). Reactions of 8a with N-benzoylacetamide (103) and N-(methoxycarbonyl) benzamide (104) afforded the1,2-dihydro-3H-pyrrol-3-ones consistent with reaction of the N-acylamidine from the external nitrogen as shown in Scheme 27. Calculations show that the coefficients of the HOMO are principally on the two nitrogens and largest on the external imino group in tautomer B (Figure 5 and Scheme 28). Thus, it appears that the reaction is kinetically favored for the tautomeric forms of 103 and 104 and frontier orbital controlled. A slow and irreversible attack at the conjugate position of 8a followed by an electrocyclic five-membered ring formation results in the regiochemistry observed for compounds 105 and 106. Thus while acyclic enaminones react mainly through C_a, the "aza-enaminones" react through the external nitrogen, consistent with a frontier-orbital controlled reaction. Methylphenylcyclopropenone and isopropylphenylcyclopropenone did not react with 103 and 104 under the reaction conditions employed, consistent with the higher energy LUMO's found for these systems when compared to 8a. Therefore, all the results obtained in this study can be understood using a frontier-orbital approach.⁶⁴

1.5.3.1 Reaction of a-TBPK, a derivative of 8a

a-(Triphenylphosporanylidene)-benzylphenylketene (a-TBPK) is a derivative of 8a prepared by its reaction with triphenylphosphine. Since this solid is stable only under anhydrous conditions it is generated in situ. The a-TBPK thus formed was reacted with 2-amino-4-methylpyridine 74b, pyridinium N-imine 92b and N-benzoylacetamidine 103, compounds that had already been reacted with the parent diphenylcyclopropenone (8a). The products formed are presented in Scheme 29 In each case triphenylphosphine was also recovered.⁶⁵ From a mechanistic viewpoint, the formation of 107-111 may be visualized as occurring through attack of the nitrogen nucleophile at the electrophilic carbon of the ketene portion of a-TBPK, followed by triphenylphosphine elimination and proton transfer (Scheme 30). The presence of this good leaving group changes the direction of the reaction when compared to the reaction of the parent 8a as shown in Scheme 24. Reaction with diethyl azodicarboxylate 112, which does not have a transferable hydrogen, formed N-acylcarbamate 113 in excellent yield together with triphenylphosphine oxide. The formation of 113 provides a new route to polysubstituted N-acylcarbamates which are versatile intermediates in the synthesis of nucleoside analogues. Their preparation is not a trivial task.⁶⁶

The above studies show elimination of the phosphine group in the final product. Earlier, in a study involving the reaction of a-TBPK with 3,3-dimethyl-2-phenyl-1-azirine, it was found that oxaphospholene 114 was formed, representing a case in which there was no loss of the phosphine group. Analogous results were obtained with 8b as a starting material as seen in Scheme 31.⁶⁷

1.5.3.2 Reactions of tosyl analog of 8a

While considerable attention had been given to the study of the reactions of diphenylcyclopropenone with pyridinium N-amines and pyridinium methylides, the chemistry of the N-tosyl analogue had been ignored. Thus a study was undertaken with the objective of comparing the reactivity of N-(p-tolylsulfonyl)diphenylcyclopropenimine 115 in reactions with mono- and disubstituted pyridinium and isoquinolinium ylides. The reaction of 115 and 116 in refluxing benzene (no reaction at room temperature) afforded 117 as the major product and 118 as the secondary product. In Scheme 32, the formation of 118 is depicted as involving a second equivalent of 116 on an initally formed ketenimine intermediate, with migration of a carboethoxy group. A heated toluene solution of 118 afforded 116 (19%), a uracil derivative 119 (28%), and the pyrimidinone 121 (41%), the latter two being other products isolated from the original reaction of 115 with 116. The thermal behavior of 118 suggests a reversibility for this process, with possible participation of a thermally unstable iminooxazine 120, in the formation of 119 and 121.⁶⁸

While no reaction occurred between 115 and pyridinium dicyanomethylide, a smooth reaction was observed with the cyanocarboethoxy analogues 122a-c producing internal salts 123a-c in high yields, which upon hydrolysis formed 124 (Scheme 33). The reaction of the dicarbethoxymethylide 122d and 115 gave a product 125 (Scheme 34), reminiscent of the product⁶¹ of monosubstituted methylides and 8a (as seen in Scheme 24). The products obtained from the reactions of 115 with pyridinium ylides were accounted for by considering pathways involving ketenimine intermediates depicted in Schemes 33 and 34. While the ketene counterparts suggested in the reactions of 8a with certain pyridinium ylides show a tendency to undergo cyclization, these results with 115 demonstrate that further reaction with nucleophiles present in the medium would be expected in this case.⁶⁸

Reaction of 115 with isoquinolinium methylides 126a-c paralleled that of pyridinium analogues, producing 127, 125a and 125c, respectively (Scheme 35). The isoquinoline N-amide derivatives 128 however, differed markedly in their reactivity toward 115. With these systems, high yields of cycloadducts 129a,b were obtained (Scheme 36). Cycloaddition products were obtained when 8a was reacted with 128a,b. This result contrasts with the reaction of diphenylcyclopropenone 8a with pyridinium N-amines where the reaction is nucleophilic (Scheme 24).⁶⁸ These results show that cyclopropenimine 115 is capable of differentiating nucleophilic and dipolar character in a homologous series of cycloiminium methylides and imines.

The reactions of pyridinium N-imines with 8a formed products consistent with a pathway involving initial nucleophilic attack on the cyclopropenone ring with subsequent ellimination of pyridine (Scheme 24). A similar behavior was seen in the reaction of 115 upon reaction with pyridinium N-carboethoxyimine 116 (Scheme 32). On the other hand, cycloaddition products 129 were obtained with isoquinoline N-amide derivatives 128. These results prompted the study of the reactions of 115 with pyridinium N-imines 92. The reaction of 115 with 92a yielded the cycloadduct 130a. The reaction of 92b in methylene chloride afforded the stable intermediate 131 as the major product. When 92c was used, the stable intermediate 131 was isolated, along with 130c. The formation of 131 provides direct evidence for the participation of 132. However, since a 1,2-dihydro intermediate was isolated in one case only, an alternative pathway to other 130 products involving inverse hydrogen transfer to 132 or a stepwise process with participation of 133 to produce a 4,4a-dihydro intermediate 134 cannot be excluded (Scheme 37).⁶⁹

2. 1-Azirines

While the behavior of simple 1-azirines toward a variety of nucleophilic reagents has received considerable attention in the literature, that of such systems containing additional electrophilic centers attached to the 3-position has been largely ignored. One example involving 1-azirine-3-carboxamides 135 with hydrazine was suggested to proceed through bicyclic aziridine intermediates with subsequent C-C or C-N bond cleavage to afford tetrahydro1,2,4-triazin-6-ones or unstable 4-amino-2-pyrazol-5-ones, depending on the nature of the 3-substituent (equation 3).⁷⁰ Although this result would suggest a promising role for azirine systems containing three electrophilic centers in the preparation of new heterocycles, no studies demonstating this potential had appeared by 1990. Therefore, studies of 1-azirines 136a and 136b with hydrazines and amidines were undertaken. Reaction of 136a with hydrazine in methanol afforded the bicyclic system 137. Similarly, 136b produced 137b (Table 6), the reaction pathway being unaffected by the presence of a substituent other than hydrogen at the 3-position. The process was represented as involving intramolecular interception of a C-N cleavage product derived from a bicyclic aziridine intermediate analogous to that depicted in equation 3. The absence of unsaturation adjacent to the aziridine portion of this intermediate should disfavor C-C bond cleavage. To address the question of initial nucleophilic attack at C-2 versus a Michael type at the 3-substituent, the behavior of 136a with substituted hydrazines was examined. The observed formation of 137c (Table 6) suggested initial attack at C-2. In an attempt to incorporate an additional carbon atom into the bicyclic products, reactions of 136 with formamidine and guanidine were studied (Scheme 38). In methanol only complex mixtures were obtained. However, in dimethyl sulfoxide, it was possible to observe the selective formation of imidazole 138, pyrimidones 139a-b or amino-s-triazine 140. The reaction pathway reflects a strong directive effect of the 3-substituent on the azirine ring. This effect was explained by considering that the reaction occurs from different tautomers (Scheme 38) resulting from C-C bond cleavage in aziridine intermediate 141, wherein product formation is thermodynamically controlled. These results represent the first example of a solvent-controlled C-N versus C-C bond cleavage process in the reactions of azirine derivatives.⁷¹

After two days, the reaction of imidazole, another nucleophile studied, with 1-azirine-3-methyl acrylate 136a (methylene chloride, room temperature) demonstrated the total consumption of reagents with formation of two new systems, one of which disappeared after an additional five days in solution. The final product obtained was 2-aza-1,3-diene 142a in 45% yield, obtained as a 10:1 mixture of Z/E isomers after column chromatography. The participation of a highly reactive aziridine intermediate 143 as the second system in the two day reaction was suggested by the NMR spectrum of the crude reaction product. Reaction of the more hindered 136b with imidazole afforded 142b (60%) as a 1:12 mixture of the Z/E isomers. Interestingly, in this case, the E-isomer predominates (Scheme 39). The reactions of these and other nucleophiles improved with the presence of catalytic amounts of the non-nucleophilic base sodium carbonate which shows that the ring opening of the aziridine intermediate is base catalyzed. The most interesting aspect of the chemistry of azadiene systems has been the constuction of heterocycles through the use of the hetero Diels-Alder reaction. The reaction of 142b with the electron deficient diethylazodicarboxylate (DEAD) and tetracyanoethylene (TCNE) formed cycloadducts 144 and 145, respectively (Scheme 39) In both cases, the diazole nucleus serves as an efficient leaving group in the initial Diels-Alder adduct. This observation calls attention to this potentially valuable feature in the azadienes 142 prepared in the present study.⁷²

Reaction of simple azirines with diphenylketene forms bicyclic aziridines by way of addition of two equivalents of ketene to the imine moiety (equation 4).⁷³

Continued interest in the effect of higher order functionality at the 3-position of the 1-azirine nucleus led to the study of the reaction of 1-azirine-3-acrylates as nucleophiles with the electrophilic diphenylketene formed from 1,2-diphenyldiazoethanone 33a. Formation of 5-pyrrolin-2-ones 146 in good yields was observed for the azirines 136b-e (Scheme 40). Alcohols added readily to the imine fragment in 146, as illustrated for the reactions of methanol with 146a-c, wherein methoxy pyrrolidinones 147a-c were obtained as mixtures of diastereomers. Under conditions of catalytic hydrogenation (H₂/Pd-C/MeOH/6H), diastereomeric mixtures of 147b-c suffered concomitant hydrogenolysis of the C₅ methoxy with intramolecular Michael addition to afford 2-azabicyclo[3.1.0]hexanes 148a-b, respectively. This reaction represents a new approach to cyclopropane ring formation.⁷⁴ When diastereomer 147b (that which contains C-4Me cis to C-5Ph) was submitted to this condition (catalytic hydrogenation), a smooth reduction to 149 was observed without any indication of hydrogenolysis. Under acidic conditions 149 forms 150 exclusively, most likely through initial loss of methanol, with no sign of an intramolecular aldol process. Extensive substitution at C-4 and C-5 in 149 appears to be responsible for this observation.

While 1-azirines react with diphenylketene according to equation 4, simple imines afford 2-azetidinones upon reaction with diphenylketene (equation 5).⁷⁴ The ready availability of 2-formyl-2H-azirine-N-aryl-imines 151, wherein both imine types are encountered, prompted the investigation of the behavior of 151 in the above reaction. Compounds 151a-c reacted with diphenylketene to form 2H-2-azirinyl-2-azetidinones 152a-c and some N-aryldiphenylacetamides 153a-c (Scheme 41). Formation of N-aryldiphenylacetamides 153a-c may be attributed to the interception by intermediate 154 of trace quantities of water in the reaction medium, in competition with ring closure to 152. The results demonstrate the enhanced reactivity of the exo-imine group in 151, furnishing a convenient route to the previously unknown 2H-2-azetidinone system in a highly diastereoselective fashion.⁷⁵

All the above reactions involve azirines with electron withdrawing groups on the 3-position (referred to as the "3-Z mode"). The next study focuses on an electron-rich substituent at the 3-position (referred to as the "3-X mode") which exerts a pushing effect on the three membered ring. To the extent that this process results in C-N bond cleavage, the resulting imino anion might be induced to undergo reaction with electrophiles.The 3-acetate derivative 155 was chosen as a simple precursor to a 3-X substituted 1-azirine. Mild base-catalyzed reaction of 155 with aldehydes and ketones is reported here. While no reaction occurred between methyl 2-phenyl-1-azirine-3-acetate 155 and a series of aldehydes and ketones used as solvents in the absence of base, the presence of DABCO in the reaction medium resulted in a smooth conversion to 3-oxazolines 156, as a 1:1 mixture of diastereomers (where these exist). Formation of 156 is consistent with participation of the intermediate 3-X substituted 1-azirine 157 generated in low concentration through proton extraction from 155 by DABCO with ring opening to 158 (Scheme 42). A nucleophilic role for DABCO in this reaction, which would invoke the participation of aziridine intermediates, was discarded inasmuch as potassium carbonate also produces 156, albeit at a slower rate. In contrast to the well-known 2-oxazoline isomers, 3-oxazolines are a rare class of compounds and only a few methods have been elaborated for their preparation.⁷⁶ Of these methods, one approach is extremely pertinent to the present discussion.⁷⁶ Aldehydes and ketones react with nitrile ylides produced from 1-azirines by photochemically promoted C-C bond cleavage (equation 6). In this case, the carbonyl moiety becomes the 1,5-fragment in the resulting 3-oxazoline. In the present study, the 1,2-fragment in 156 originates from the carbonyl component. The two methodolgies are therefore complementary, although it should be noted that acceptable yields of 3-oxazolines in the photochemical process require the presence of electron-withdrawing substituents on the carbonyl moiety. The mild reaction conditions employed in the present study permit, for the first time, construction of a 3-oxazoline nucleus containing versatile functional groups at the 2-position (i.e. styryl (156e) and furyl (156f)).⁷⁷

3. Retrospective

This account represents a good part of our published work with enaminones, diazocarbonyls and small rings. There is a considerable amount of unpublished data which was not included here, nor were reactions of diazocarbonyl related compounds with nucleophiles⁷⁸ other than enaminones. Basically we started with the concepts of nucleophiles reacting with electrophiles using systems that are ambident and/or ambiphilic or can form interesting intermediates with these properties, and used structure considerations to generate chemical possibilities. A mechanistic approach was used to understand the reactivity. The addition of the more subtle notions of molecular orbital theory, principally the concept that frontier orbitals are important to reactivity, came when some of the experimental data had already been obtained. Having found that the reactions are, in general, frontier-orbital controlled, this additional concept added a new dimension to our studies, allowing us to make more well founded suggestions about reaction mechanisms. For example, as suggested by frontier-orbital treatment, acyclic enaminones tend to react at the C_a position and diphenylcyclopropenone reacts by conjugate addition. Also, molecular orbital calculations suggest that the Me-C of methyl phenylcyclopropenone (8b) should be the kinetically favored position for reaction with nucleophiles.

These studies have shown that the reactions of diphenylcyclopropenone seem to involve ketene intermediates formed from ring opening of primary adducts of nucleophilic attack. Reactions with pyridinium N-imines, 8b show more tendency toward cycloaddition than does diphenylcyclopropenone. Also, reactions of 8a and diphenylcyclopropenyl tosylimine (115) with isoquinoline N-amide derivatives are more likely to form cycloadducts than with pyridinium N-amide derivatives.

As already mentioned, many of the chemical systems obtained by us in these and other works have medicinal and biological possibilities. Besides preparing compounds with possible biological activity, recently there has appeared in the literature examples of enaminones, azirines and cyclopropenones (our starting reagents) with medicinal activity. Thus, methyl 4-(p-chlorophenylamino)-6-methyl-2-oxocyclohex-3-ene-1-oate and methyl 4-(benzylamino)-6-methyl-2-oxocyclohex-3-ene-1-oate are two anticonvulsant enaminones.⁷⁹ Azirinomicine, the first naturally occurring 1-azirine (isolated from the Steptomyces aureaus bacteria) is an antibiotic⁸⁰ and dysidazirine, the first example of this strained ring class of heterocycles from a marine source, has citotoxic activity.⁸⁰ The cyclopropenone penitricine has antibiotic activity and has been used as a model for other biologically active cyclopropenones.⁸¹

The chemical understanding obtained from these and other studies should help researchers understand the mode of action of these and related biologically active compounds and help produce more effective drugs. With this, we can see that the continuation of this research along the medicinal line presents itself as an exciting possibility for the future.

References

1. Greenhill, J. V.; Chem. Soc. Rev. 1977, 6, 277.

2. Lue, P.; Greenhill, J. V.; Advances in Heterocyclic Chemistry, 1997, 67, 207.

3. Grob, C. A.; Wilkens, H. J.; Helv. Chim. Acta 1967, 725; Dudek, G. O.; Holm, R. H.; J. Am. Chem. Soc. 1962, 84, 2691; Dudek, G. O.; Volpp, G.; J. Am. Chem. Soc. 1963, 85, 2697; Dudek, G. O.; Dudek, E.P.; J. Am. Chem. Soc. 1964, 86, 4238; Dabrowski, J.; Spectrochim. Acta 1963, 19, 475; Weinstein, J.; Wyman, G.; J. Org. Chem. 1958, 23 , 1618; Collman, J.; Alive, R.; Sohar, P.; J. Am. Chem. Soc. 1958, 80, 1100; Dudek, G. O.; J. Org. Chem. 1965, 30, 548; Brown, N. M. D.; Nonhebel, D. C.; Tetrahedron 1968, 24, 5655; Célérier, J.-P.; Lhommet, G.; J. Heterocycl. Chem. 1982, 19, 481.

4. Greenhill, J. V.; J. Chem. Soc., Ser. B 1969, 299.

5. Eberlin, M. N.; Takahata, Y.; Kascheres, C.; J. Mol. Struct (Theochem) 1990, 207, 143.

6. Smith, D.; Taylor, P. J.; Spectrochim. Acta, Part A 1976, 32, 1477; Kashima, C.; Aoyama, H.; Yamamoto, Y.; Nishio, T.; Yamada, K.; J. Chem. Soc., Perkin Trans 2 1975, 665; Gomez-Sánches, A.; Aldave, M. T.; Scheidegger, U.; J. Chem. Soc., Ser. C 1968, 2570; Gomez-Sánches, A.; Valle, A. M.; Bellanato, J.; J. Chem. Soc., Perkin Trans 2 1973, 15; Kozerski, L.; von Philipsborn, W.; Org. Magn. Reson. 1981, 17, 301; Dabrowski, J.; Kamienska-Trela, K.; J. Am. Chem. Soc. 1976, 98, 2826; Ostercamp, D. L.; J. Org. Chem. 1970, 35, 1632; Kozerski, L.; Kamienska-Trela, K.; Kania, L.; von Philipsborn, W.; Helv. Chem. Acta 1983, 66, 2113.

7. Bilinski, V.; Steinfels, M. A.; Dreiding, A. S.; Helv. Chem. Acta 1972, 55, 1065.

8. Eicher, T.; Bohn, S.; Tetrahedron Lett. 1972, 2603.

9. Kascheres, C.; Kascheres, A.; Pilli, P. S. H.; J. Org. Chem. 1980, 45, 5340.

10. Fleming, I.; Frontier Orbitals and Organic Chemical Reactions, Wiley: Chichester, 1976.

11. Kascheres, A.; Kascheres, C.; Herrera Braga, A. C.; J. Org. Chem. 1993, 58, 1702.

12. Robins, D. J.; Fortshr. Chem. Org. Naturst. 1981, 41, 115 and references cited therein; Atal, C. K.; Lloydia 1978, 41, 312.

13. Yechezkel, T.; Ghera, E.; Ostercamp, D.; Hassner, A.; J. Org. Chem. 1995, 60, 5235; Keck, G. E.; Nickell, D. G.; J. Am. Chem. Soc. 1980, 102, 3632; Robins, D. J.; Nat. Prod. Rep. 1991, 8, 213; Daí, W. _M.; Naggao, Y.; Heterocycles 1990, 30, 1231; Robins, D. J.; Nat. Prod. Rep. 1990, 7. 377; Pinder, A.. R.; Nat. Prod. Rep. 1990, 7, 447.

14. Chamberlin, A. R.; Chung, J. Y. L.; J. Org. Chem. 1985, 50, 4425.

15. Staskun, B.; Van Es, T.; J. Chem. Soc. (C) 1966, 531.

16. Kascheres, A.; Rodriques, R. A.. F. ; Tetrahedron 1996, 52, 12919.

17. Satake, K.; Kumugai, T.; Mukai, T. Chem. Lett. 1983, 5, 743; Roeber, H..; Matusch, R.; Hartke, K.; Chem. Ber. 1975, 108, 3247; Mukai, T.; Kumagai, T.; Tanaka, S.; Jpn. Kokai Tokkyo Koho JP 62,207,275 [87,207,275] 1988 (CA 108: 186728v); Soberg, Y. ; Z. Naturforsh, C.; Bioscience 1977, 32, 292.

18. Arita, M.; Saito, T.; Fukumasu, Y.; Haga, K.; Jpn. Kokai Tokkyo Koho Jp 04,270,283 [92,270,283] 1993 (CA 118: 191719d).

19. Pulat, M.; Greif, D.; Kleinpeter, E. Z.; Chem. 1988, 28, 345; Baruah, P. D.; Mukherjee, S.; Mahajan, M. P.; Tetrahedron 1990, 46, 1951.

20. Ciabattoni, J.; Berchtold, G. A.; J. Org. Chem. 1966 31, 1336.

21. Kascheres, A.; Kascheres, C.; Herrera Braga, A. C.; J. Org. Chem. 1993, 58, 1702.

22. Eberlin, M. N.; Takahata, Y.; Kascheres, C.; J. Org. Chem. 1990, 55, 5150.

23. Asbrink, L.; Fridh, C.; Lindholm, E.; Chem. Phys. Lett. 1977, 53, 63; Ibid. 1977, 52, 69; Ibid. 1977, 52, 72; Asbrink, L.; Fridh, C.; Lindholm, E. de Bruijn, S.; Chem. Phys. Lett., 1979, 66, 411; Lindholm, E..; Asbrink, L.; Molecular Orbitals and Their Energies, Studies by the Semiempirical HAM Method, Lecture Notes in Chemistry, Springer-Verlag: Berllin, 1985, vol. 38.

24. Eberlin, M. N.; Kascheres C.; J. Org. Chem., 1988, 53, 2084.

25. Eberlin, M. N.; PhD Thesis, Universidade Estadual de Campinas, Campinas 1988.

26. Augusti, R.; Eberlin, M. N.; Kascheres, C.; J. Heterocyclic. Chem. 1995, 32, 1355.

27. Schlunegger, U. P; Angew. Chem. Internat. Ed. Engl. 1975, 14, 679; Cooks, R. G.; Beynon, J. H.; Caprioli, R. M.; Lester, G. R.; Metastable Ions, Elsevier Scientific Publishing Company: Amsterdam, 1973, p. 42-44; Beynon, J. H.; Cooks, R. G.; Amy, J. W.; Baitinger, W. E.; Ridley, T. Y.; Anal. Chem. 1973, 45, 1023A.

28. Eberlin, M. N.; MSci. Dissertation, Universidade Estadual de Campinas, Campinas, 1984.

29. Augusti , R.; Kascheres, C.; J. Org. Chem. 1993, 58, 7079.

30. Vogt, B. R.; U.S. patent 4,112,098, 1978 (CA 92: 146801).

31. Kohn, E. C.; Liotta, A.; J. Natl. Cancer Inst. 1990, 82, 54.

32. D. A.; Settimo, A.; Livi, O.; Biagi, G.; Primofiore. G.; Masoni, G.; Farmaco Ed. Sci. 1982, 37, 728.

33. Nielsen, F. E.; Pedersen, E. B.; Begtrip, M.; Liebigs. Ann. Chem. 1984, 11, 1848.

34. Buckle, D. R.; Smith, H.; Eur. Patent Appl. EP34,004, 1982 (CA 96: 35265z); D. R. Buckle, Rockell, C. J. M.; J. Chem. Soc., Perkin Trans. 1 1982, 627; Buckle, D. R.; Outred, D. J.; C. J. M. Rockell, Smith, H.; Spicer, B. A.; J. Med. Chem. 1983, 26, 251.

35. Meier, R;. Eur. Patent Appl. EP229,011, 1987 (CA 107: 198336h).

36. Biagi, G.; Ferreti, M.; Livi, O.; Scartoni, V.; Lucacchini, A.; Mazzoni, M.; Farmaco Ed. Sci. 1986, 41, 388.

37. Augusti, R.; Kascheres, C.; Tetrahedron 1994, 50, 6723 and references cited therein.

38. Kazitsyna, L. A.; Kikot, B. S.; Vpadysheva, A. V.; Russ. Chem. Rev. 1966, 35, 388.

39. Kraus, G. A.; Nagy, J. O.; DeLano, J.; Tetrahedron 1985, 41, 2337.

40. Khaskin, B. A.; Molodova, O.D.; Torgasheva, N. A.; Russ. Chem. Rev. 1992, 61, 3.

41. Kazitsyna, L. A.; Klyueva, N. D.; Romanova, K. V.; Dokl. Chem. Roc. Acad. Sci. USSR 1968, 183, 955; Kraaijeverld, A.; Having, E.; Rec. Trav. Chem. 1954, 73, 537.

42. Figueiredo, L. J. O.; Kascheres, C.; J. Org. Chem. 1997, 62, 1164.

43. Rodrigues, B. L.; Gambardella, M. T. do P.; Figueiredo, L. J. O.; Kascheres, C.; Acta Cryst. 1996, C52, 705.

44. Chemia, D. S.; Zyss, J.; Nonlinear Optical Properties of Organic Molecules and Crystals, Academic Press: New York, 1987; Williams, D. J.; Nonlinear Optical Properties of Organic and Polimeric Materials, ACS Symp. Ser 1983, 233.

45. DeQuan, L.; Ratner, M. A.; Marks, J. J.; J. Am. Chem. Soc. 1988, 110, 1707.

46. Kanis, D. R.; Ratner, M. A.; Marks, T. J.; Chem. Rev. 1994, 94, 195.

47. Ulman, A.; Willand, C. S.; Kohler, W.; Robello, D. R.; Williams, D. J.; Handlesy, L.; J. Am. Chem. Soc. 1990, 112, 7083.

48. Kieckononowicz, K. K.; Ligneau, X.; Schwartz, J.-C.; Schunack, W.; Arch. Pharm. (Weinheim Ger.) 1995, 328, 469; Haunert, F.; Bolli, M. H.; Hinzen, B., Ley, S. V.; J. Chem. Soc., Perkin Trans. 1 1998, 15, 2235; Aberola, A.; Gonzalez-Ortega, A.; Sadaba, M. L.; Sanudo, M. C.; J. Chem. Soc. Perkin Trans. 1 1998, 24, 4061; Oikawa, N.; Mueller, C.; Kunz, M. W.; Lichtenthaler, F. W.; Carbohydrate. Res. 1998, 309, 269; Makino, K.; Kim, H. S.; Kurasawa, Y.; J. Heterocyclic Chem 1998, 35, 489; Krygowski, T. M.; Anulewicz, R.; Cyranski, M. K.; Puchala, A.; Rasala, D.; Tetrahedron 1998, 54, 12295.

49. Kascheres, C.; Negri, G.; Gambardella, M. T. P.; Santos, R. H. A.; J. Braz. Chem. Soc. 1994, 5, 31.

50. Negri, G.; Kascheres, C.; J. Heterocyclic Chem. 2001, 38, 109.

51. Kascheres, C.; Negri, G.;. Ferreira, M. M. C.; Sabino, L. C.; J. Chem. Soc., Perkin Trans. 2 2001, 2237.

52. Hassner, A.; Kascheres, A.; J. Org. Chem. 1972, 37, 2328.

53. Lown, J. W.; Matsumoto, K.; Can. J. Chem. 1971, 49, 1971.

54. Kascheres, A.; Rodrigues, J. A. R.; J. Org. Chem 1975, 40, 1440.

55. Eicher, T.; Franke, G.; Liebigs Ann. Chem. 1981, 1337.

56. Kascheres, A.; Kascheres, C.; Rodrigues, J. A. R.; Synth. Commun. 1984, 14, 905.

57. Kascheres, A.; Reyes R.,J. L.; Fonseca, S.M.; Heterocycles 1984, 22, 2529.

58. Kascheres, A.; Kascheres, C.; Rodrigues, J.A.R.; Santana, A. R. A.; J. Org. Chem. 1976, 41, 3546.

59. Kascheres, A.; Correio Filho, J.; Cunha, S.; Tetrahedron 1993, 49, 381.

60. Lown, J. W.; Matsumoto, K.; Can. J. Chem. 1972, 50, 584; Sasaki, T.; Kanematsu, K.; Kakeki, A.; J. Org. Chem. 1971, 36, 2451

61. Kascheres, A; Marchi Jr., D.; J. Org. Chem. 1975, 40, 2985.

62. Kascheres, A.; Marchi Jr., D.; J. Chem. Soc., Chem Comm. 1976, 275.

63. Kascheres, A.; Marchi,Jr., D.; Rodrigues, J. A. R.; J. Org. Chem. 1978, 43, 2892.

64. Cunha, S.; Kascheres, A.; J. Braz. Chem. Soc. 2001, 12, 481.

65. Cunha, S.; Kascheres, A.; J. Braz. Chem. Soc. 2002, 13, 687.

66. Westwook, N. B.; Walker, R. T.; Tetrahedron 1998, 54, 13391; Wyatt, P. G.; Anslow, A. S.;Coomber, B. A.; Cousins, R. P. C.; Evans, D. N.; Gilbert, V. S.; Humber, D. C.; Patemoster, I. L.; Sollins, S. I.; Tapolezay, D. J.; Weingarten, G. G.; Nucleosides Nucleotides 1995, 14, 2039.

67. Kascheres, A.; Joussef, A. C.; Duarte, H. C.; Tetrahedron Lett. 1983, 24, 1837.

68. Pilli, R. A.; Rodrigues, J. A. R.; Kascheres, A.; J. Org. Chem. 1983, 48, 1084.

69. Kascheres, A.; Rodrigues, J. A. R.; Pilli, R. A.; Heterocycles 1984, 22, 1709.

70. Nishiwaki, T.; Saito, T.; J. Chem. Soc. C 1971, 2648.

71. Kascheres, A.; Oliveira, C. M. A.; de Azevedo, M. B. M.; Nobre, C. M. S.; J. Org. Chem. 1991, 56, 7.

72. Barboso, M. T.; Kascheres, A.; J. Org. Chem. 1999, 64, 49.

73. Hassner, A.; Miller, A. S.; Haddadin, M. J;. Tetrahedron Lett. 1972, 1353; Woerner, F. R.; Reimlinger, H.; Merenyi, R.; Chem. Ber. 1971, 104, 2786.

74. Kascheres, A.; Nunes Jr., J.; Brandão, F.; Tetrahedron 1997, 53, 7089.

75. Nunes Jr., J.; Kascheres, A.; J. Heterocyclic Chem. 1997, 34, 341.

76. Huisgen, R.; Rash, R.; Runge, K.; Stangl, H.; Chem. Ber. 1972, 105, 1279; For a review see: Nair, V. Azirines. In The Chemistry of Heterocyclic Compounds; Hassner, A., ed., Interscience: New York, 1983, vol 42 (Small Ring Heterocycles, Pt. 1), p. 215; Pfoertner, K. H.; Bernauer, K.; Kaufmann, F.; Lorch, E.; Helv. Chim. Acta 1985, 68, 584; Hassner, A.; Amarasekara, A. S.; Andisik, D. J.; J. Org. Chem. 1988, 53, 27; Singh, G. S.; J. Indian Chem. 1987, 26B, 270.

77. Sá, M. C. M.; Kascheres, A.; J. Org. Chem. 1996, 61, 3749.

78. Sparrapan, R.; Gambardella, M. T. P.; Kascheres, C.; J. Org. Chem. 1995, 60, 3975.

79. Scott, K. R.; Edafiogho, I. O.; Richardson, E. L..; Ferrar, V. A.; Moore, J. A. Tietz, E. I.; Hinko, C. N. Chang, H.; El-Assadi, A.; Nicholson, J. M.; J. Med. Chem. 1993, 36, ; Scott, K. R.; Moore, J. A.; Sims, L. D.'Tonnu, A. D.; J. Med. Chem. 1995, 38, 4033.

80. Miller, T. W.;Tristam, E. W.; Wolf, F. J.; J. Antibiotic 1971, 24, 48; Ireland, C. M.; Molinsky, T. F.; J. Org. Chem. 1988, 53, 2103.

81. Okuda, T.; Yokose, K.; Furumai, T.; Maruyama, H. B.; J. Antibiotics 1984, 37, 718; Ando, R.; Morinaka, Y.; Tokuyama, H.; Isaka, M.; Nakamura, E.; J. Am. Chem. Soc. 1993, 115 1174.

Received: December 2, 2002

Published on the web: December 9, 2003

FAPESP helped in meeting the publication costs of this article

I dedicate this work to Al Kascheres (with whom I have the joy of sharing this wonderful life) on the occasion of his 60^th birthday (May 24, 2003)

Publication Dates

• Publication in this collection
  11 Feb 2004
• Date of issue
  Dec 2003

History

• Received
  02 Dec 2002