SciELO - Scientific Electronic Library Online

 
vol.46 issue3CONFORMATIONAL DYNAMICS OF SUBSTITUTED N-ACETYL-N-PHENYLBENZYLAMINES. 1H-DNMR AND AM1-MO STUDYOPTIMIZACION DE UNA TECNICA DE MUESTREO IN SITU: ANÁLISIS DE Cu Y Fe EN SISTEMAS ACUÁTICOS author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

Share


Boletín de la Sociedad Chilena de Química

Print version ISSN 0366-1644

Bol. Soc. Chil. Quím. vol.46 n.3 Concepción Sept. 2001

http://dx.doi.org/10.4067/S0366-16442001000300006 

A THEORETICAL STUDY OF THE STEREOCHEMICAL COURSE
OF THE ACID CYCLIZATION OF TWO 5-OXOGERMACREN-
6,12-OLIDES EPIMERIC IN C4

VICTOR KESTERNICH1*, CARLOS CONTRERAS-ORTEGA1 and
ROLANDO MARTINEZ2

1)Departamento de Química, Universidad Católica del Norte, Casilla 1280, Av. Angamos
0610, Antofagasta, Chile. e-mail: vkestern@ucn.cl, Fax: (56-55)355632

2)Instituto de Química, Universidad Austral de Chile, Casilla 567, Valdivia, Chile.

(Received: July 6, 2000 - Accepted: May 4, 2001)

SUMMARY

The theoretical study of both the conformational distribution and the acid cyclization of two epimeric 5-oxogermacren-6,12-olides in C4 is described. Our results on the stereochemistry of the cyclization products, 5-hydroxyguaian-6,12-olides, agree with experimental results reported previously. Curtin-Hammet principle and Hammond postulate are used to propose reaction pathways consistent with above results. The theoretical results were obtained by using a MM2 program from HyperChem 4.5.

KEYWORDS: Sesquiterpenic lactones, germacranolides, guaianolides, MM2 program.

RESUMEN

Se ha realizado el análisis conformacional y el estudio teórico del curso estereoquímico de las ciclaciones ácidas de dos germacranolidas epiméricas en C4, empleando el programa MM2 de HyperChem 4.5. La esteroquímica deducida de los productos de reacción, 5-hydroxyguaian-6,12-olides, son plenamente concordante con los resultados experimentales reportados previamente. Tanto el principio de Curtin-Hammet como el postulado de Hammond, permiten proponer una explicación a los resultados obtenidos.

PALABRAS CLAVES: lactonas, sesquiterpénicas, germacranótidas, guayanolidas, programa MM2.

Germacranolides and guaianolides have been studied both from the structural and reactivity viewpoints1-6, due to their importance as chemical models to study guaianolide and germacranolide - like intermediates occurring in reactions that take place in the biological systems of some plants. 5-oxogermacren-6,12-olides and 5-hydroxyguaian-6,12-olides (Figure 1) are compounds belonging to the sesquiterpenic lactone group whose lactonic ring is commonly cycled at C6 or C8, with a mainly trans stereochemistry.

Fig.1. Structure of 5-oxogermacrenolides 1 and 2, and 5-hydroxyguaianolides 3-8.

This paper shows the results of a theoretical study to elucidate the acid cycling stereochemistry of two 5-oxogermacren-6,12-olides methyl-epimeric on C4, 1 and 2, which lead to obtaining 5-hydroxyguaian-6,12-olides and whose synthesis stereochemistry and acidic cyclization have been reported previously3,4. Germacranolides can generate different types of compounds depending upon several factors: (a) the stereochemistry of their sustituients; (b) the position of these sustituent; and (c) the funtional groups present in them. To our knowledge, using of semi-empirical methods for the studying of the acid cyclation of several germacranolides to form guaianolides have been reported elsewhere7, but not for the type of guaianolides we here present. Because of the agreement of our theoretical results with those experimentally obtained, the methodology and conclusions of this work seem useful as a contribution for predicting the stereochemical course of these type of reactions from a theoretical point of view.

RESULTS AND DISCUSSION

Theoretical studies via molecular mechanics were performed by using the MM2 method from the Hyperchem8 program. Others calculation methods could also be used, but the decision about which is the most appropiate one seems to be a rather subjective matter, as it is recognized by the scientific community familiar with calculations oriented to our purposes9. In that regard, the MM2 method is so valid as many others, but it showed to be useful for the goal of estimating the course of the acidic cyclation we are concerned with. No doubts, for those interested in refining numerical parameters, which is not our case, other methods migh be more convenients.

The theoretical conformational study of 1 was initiated by calculating the potential surfaces generated by the movement of fragments C1-C2-C3-C4 and C7-C8-C9-C10 of the molecule, representing the fragment movement of each of the relative dispositions of the carbonyl in C5 and the methyl in C10 (Table I). These correspond with two syn positions of these groups (a,a) and (b,b) and two anti positions (a,b) and (b,a). A total of 17 possible conformers designated from A to Q were obtained from the calculation of these surfaces. Table II shows the relative energy and population of each conformer with respect to the most stable, that is, A. Only 6 conformers (A, B, C, D, M and N) have a significant population and two of them (A and B) amount to a total population of 89.3 %. Therefore, ketone 1 should be mainly in conformational equilibrium between conformers A and B at a ratio of 4.8:1. A is a conformation CC (chair-chair) and B a conformation TC (twist-chair) (Figure 2). In both conformers, the C5 carbonyl and H1 have a syn disposition on the alpha face of the molecule. The lower stability of B (930 cal/mol) relative to A, might be due to the stronger interactions present in B between the C10-Me and the C5-carbonyl, a fact which could give account of the much lower population of B.

Table I. Relatives dispositions of carbonyl at C5 and methyl at C10- ketone 1.


C5-Carbonyl

C10-Methyl


a

a

a

b

b

a

b

b


Table II. Relative energies and populations of conformers of ketone 1 from MM2 calculations


Conformer

Relative energy
kcal/mol

Population %


A

0.0

74.02

B

0.93

15.3

C

3.01

0.5

D

4.04

0.1

E

6.92

-

F

7.70

-

G

8.13-

H

9.27

-

I

8.02

-

J

8.09

-

K

10.990

-

L

10.040

-

M

1.78

3.7

N

1.47

6.2

O

5.67

-

P

5.74

-

Q

4.48

-


-

Fig. 2. Conformers A (CC) and B (TC) for compound 1.

In the theoretical conformational study of 2, by applying the same methodology as that of 1, 18 conformers named from A to Q were obtained. These have two slightly different rotations for G named G1 and G2 (Table III). The most stable conformers are A and B, amounting to a population of 84.2%, with a conformation CC and TC, respectively (Figure 3). The conformation B is more stable than the conformer A (70 cal/mol). This could be due to the absence in B of the C4Me-C10Me interactions that are present in the conformation A. Besides, a conformer M (10.8%) stands out, in which the carbonyl in C5 and H1 are anti-disposed on the alpha and beta faces, respectively.

Table III. Relative energies and populations of conformers of ketone 2 from MM2 calculations

 


Conformer

Relative energy
kcal/mol

Population
%


A

0.07

39.8

B

0.00

44.4

C

2.98

0.3

D

2.98

0.3

E

2.67

0.5

F

5.63

-

G1

4.00

-

G2

6.25

-

H

6.50

-

I

2.42

0.7

J

3.82

-

K

3.74

-

L

5.51

-

M

0.83

10.8

N

1.60

3.0

O

4.39

-

P

6.87

-

Q

4.75

-


Fig. 3. Conformers A (CC) and B (TC) for compound 2.

For the evaluation of the relative stabilities of the possible cyclization products, the fact that the bicyclical system of the guaianes studied, 3-8, is conformationally very rigid has been considered. This rigidity arises as a consequence of the fusion of two rings, A/B, the trans-lactonic ring and the double bond exocyclic D10(14). Thus, the conformational study is reduced to the calculation of a single-chair cycloheptanic conformation, containing three rotamers due to rotation of the O-H bond into the segment C1-C5-O-H.

The results in Table IV from MM2 calculations show that the most stable rotamers exhibit an anti disposition of the hydroxylic hydrogen and A/B rings fusion bonding (C1-C5) in all cases (3-8). This disposition enables the formation of a hydrogen bridge with the lactonic oxygen. Also, if the methyl group on C4 is in alpha dispositions 3-4, the most stable fusion would be cis-(1a-H,5a-OH) 4 with a global population of 99.95%, as opposed to the junction cis-(1b-H, 5b-OH) 3 that amounts only to 0.05%. On the contrary, if the methyl on C4 is in beta dispositions 5-8, the most stable fusion would be cis-(1b-H, 5b-OH) 6, with a global population of 93.53% as opposed to the junction cis-(1a-H,5a-OH) 5 that amounts only to 3.88 % and to the other two junctions, trans-a,b and trans-b-a, with even lower populations. We can then conclude that the cyclization of ketone 1 will mainly lead to guaianolide 4, while ketone 2 will mainly generate product 6. This is strongly confirmed by experiments3.

Table IV. Relative energy and population of the different rotamers for each of the possible products added

As a general criterion, it is accepted that the main cyclization product evolves from those conformers of the starting ketone having the carbonyl in C5 and H1 in the same disposition that the OH in C5 and H1 are to exhibit in the product4.

From above, the cyclization of compound 1 to give guaianolide 4 would take place through the major conformers A (CC) and B (TC). In such a case, the steric course of the cyclation of ketone 1 should be controlled by the conformations with the highest populations, i. e., by those more stable.

According to the mentioned criterion, in analyzing the cyclization reaction pathway for ketone 2 from the side of the reactants, the following conclusion is obtained: if any of its two major conformers A and B (amounting to 84.2%) cyclated, guaianolide 5 with a cis-(1a-H,5a-OH) junction should generate. In the same way, if cyclization took place through M (10.8%) (Figure 4), trans-(1b-H,5 -OH)-guaianolide 8 should generate. However, experimental results show that the cyclization product exhibits the beta disposition of 5-OH, like in structure 6, in agreement with the theoretical conclusions of this work. This fact leads us to the conclusion that the cyclization occurs from the minor conformers E to H (~1%) (Figures 5 to 9), in which H1 and the ketonic carbonyl are syn-disposed, oriented by the beta face of the molecule. Therefore, the steric course of the cyclation of ketone 2 appears controlled by the conformations with the lowest populations, i. e., by those less stable, in opposition to the case of ketone 1.In order to explain this fact, the Curtin-Hammet principle10 must be applied. This principle states that if a substrate is conformationally heterogeneous, the final product is not determined by the conformation exhibiting the highest population (the most stable), but by the one whose transition state towards the products requests the lowest activation energy, if the activation energies of the competing reactions are greater than the activation energy of the conformational interconversion. It then follows that conformers A and B interconvert into those conformers from E to H that better satisfy the energy relationships requested by the Curtin-Hammet principle. Then they evolve to the final product guaianolides 6. This means that these interconvertions occur much faster than the competing reactions. To be consistent with our analysis, we must accept that the cyclization reaction of ketone 1 occurs via its conformers A and/or B because one or both of them are the conformers which better satisfy the energy requirements of the Curtin-Hammet principle and not because their populations are the highest. We are here reasonably assuming that the relative magnitudes of the energies involved in this reaction are of the same order of magnitude as the energies involved in the reaction of ketone 2. We say that E to H are the possible precursor conformers of the product 6 in the same way as conformers A and B of ketone 1 are the possible precursor conformers of product 4.

Fig. 4. Conformer M for compound 2.

Fig. 5. Conformer E for compound 2.

Fig. 6. Conformer F for compound 2.

Fig. 7. Conformer G1 for compound 2.

Fig. 8. Conformer G2 for compound 2.

Fig. 9. Conformer H for compound 2.

Finally, the results show that the disposition of the methyl group on C4 seems to be responsible for the stereochemistry of the junction between the two rings formed in the cyclization reaction of both ketone 1 and 2. This is under study.

Calculations by MM2 enable to know the values of the so-called strain energy. These values allow seeing how energetically far from the theoretical most probable products are the theoretical most probable reactants and thus the strain energy is a useful parameter for comparing the stability of structurally different compounds and for estimating the structure of the transitions states of some reactions. This is crucial to understanding reaction mechanisms. Rotational and electronic factors have not been considered in our calculations, since it has been reported that they are not significant in this type of molecules (reference 3 and references there in). Results for some conformers of starting ketone and the corresponding possible cyclization products are shown in Tables V and VI, respectively.

Table V. Strain energies of some of the conformers of starting ketone 1 and 2 from MM2 calculations.


Compounds

1-A,

1-B

2-A

2-B

2-E

2-F

2-G1

2-G2

2-H

2M


Strain energy (kcal/mol)

10.39

11.32

11.93

11.86

14.53

17.49

15.86

18.11

18.36

12.69


Table VI. Strain energies of the conformers (rotamers) of products guaianolides from MM2 calculations. 3 and 4 are products of ketone 1 while 5-8 are products of ketone 2.


Compounds

3

4

5

6

7

8


Strain energy
(kcal/mol)

18.44

13.80

15.61

13.99

17.88

17.90


Hammond's postulate11 states: "if two states, as for example, a transition state and an unstable intermediate, occur consecutively during a reaction process and have nearly the same energy content, their interconversion will involve only a small reorganization of molecular structure. It then follows that in an exothermic step with low activation energy, the transition state will structurally resemble the reactant since they are close in energy and therefore interconverted by a small structural change. In the case of an endothermic step with low activation energy from the side of the product, the energy of the transition state is similar to that of the product and the transition state will structurally resemble the product. Finally, in the case where the step involves a transition state which is a good deal higher in energy than either the reactant or the product, neither the reactant nor the product will be a good model of the transition state.

According to our conclusions about the cyclization reaction pathways and the results in Tables 5 and 6, reactant conformers A and B of ketone 1 evolve to product guaianolide 4 through endothermic reactions (+3.4 and +2.5 kcal/mol, respectively) while reactant conformers E, F, G1, G2 and H of ketone 2 evolve to product guaianolide 6 through exothermic reactions (-0.5, -3.5, -1.9, -4.1 and -4.4 kcal/mol respectively). Let us assume that above reactions are one-step reactions, an assumption that it is well supported by the literature in the field1-3. Let also assume that the potential energy values of their respective precursor reactants, products, and transition states fit well the requirements of Hammond's postulate to properly apply it to the estimation of the transition state structures. Observe that the potential energy changes involved in both types of reactions, though opposite in signs, are of the same order of magnitude. It could then follow that the cyclization transition state in case of ketone 1, whether from precursor A or B, should be similar to a guaianolide, and, more specifically, similar to guaianolide 4. In the case of ketone 2 the transition state in all possible reaction pathways above should be similar to a ketone, and more specifically to some of the conformers from E to H. These conclusions seem useful as starting point to finally elucidate which the transition states in these reactions should be. Work in this regard is in progress.

ACKNOWLEDGEMENTS

We thank DGICT (Grant DGI/C04/98), Universidad Católica del Norte and DID, (Grant S-98-30) Universidad Austral de Chile, Valdivia.

REFERENCES

1. A.G. González, A. Galindo, J.A. Palenzuela, H. Mansilla. Tetrahedron Lett. 27, 2771 (1986)         [ Links ]

2. A. González, A. Galindo, H. Mansilla, A. Gutiérrez, J. A. Palenzuela, M.A. Gómez-Rodríguez, M. Martínez-Ripoll, S. García-Blanco. J.Org. Chem. 50, 5856 (1985).         [ Links ]

3. .A. González, A. Galindo, M.M. Afonso, H. Mansilla, M. López. Tetrahedron 44, 4585, (1988)         [ Links ]

4. L. Fitjer, A. Malich, C. Peschke, S. Kluge, R. Gerke, B. Rissom, J. Weiser, M.Noltemeyer, J. Am. Chem. Soc. 117, 9180 (1995)         [ Links ]

5. Victor Kesternich, Rolando Martínez. Bol. Soc. Chil. Quím.,42,401-403 (1997).         [ Links ]

6. V. Kesternich, L. Vera and R. Martínez. Bol.Soc. Chil. Quím. 44, 243 (1999)         [ Links ]

7. A.Galindo et al, 52,11827-40 (1996)

8. HyperChem® 4.5, Hyper Cube Inc.

9. Victor Kesternich and Rolando Martínez, Pag 38, XXIII Jornadas Chilenas de Química, 24-27 de Noviembre, 1999G.S.

10. D.Y. Curtin, Rec. Chem. Prog. 15, 11 (1954);         [ Links ] E.L. Eliel, Stereochemistry of carbon Compounds, Mc Graw-Hill, New York, 1962, pp 151-152, 237-238.         [ Links ]

11. Hammond, J. Am. Soc. 77, 334 (1955)         [ Links ]

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License