Synlett 2024; 35(20): 2441-2446
DOI: 10.1055/a-2384-6736
letter
Special Issue to Celebrate the 75th Birthday of Prof. B. C. Ranu

Design and Synthesis of Out/Out, Out/In, and In/In Epoxides in Polycyclic Cage Frameworks

Sambasivarao Kotha
a   Department of Chemistry, Indian Institute of Technology-Bombay, Powai, 400076, India
b   Department of Chemistry, Sunandan Divatia School of Science, SVKM’s NMIMS (Deemed-to-be) University, Vile Parle (West), Mumbai, 400056, India
,
Mohammad Salman
a   Department of Chemistry, Indian Institute of Technology-Bombay, Powai, 400076, India
› Author Affiliations

M.S. thank the University Grants Commission (UGC), New Delhi for the award of a research fellowship.
 


Abstract

We report a useful synthetic approach to assemble in/in epoxide, in/out epoxide, and out/out epoxide in cage systems using the Corey–Chaykovsky reaction and the Peterson olefination as key steps. In this regard, a variety of pentacycloundecane (PCUD) based cage compounds containing oxirane rings with diverse stereochemical disposition were synthesized via a simple synthetic sequence. Five cage diones were used for this purpose, and the starting cage diones were prepared with easily accessible starting materials such as 1,4-hydroquinone derivatives and cyclopentadiene. Here, we have used the Diels–Alder (DA) reaction, a [2+2] photocycloaddition, the Corey–Chaykovsky reaction, and the Peterson olefination as crucial steps to prepare the target molecules.


#

During the past 60 years, organic chemists have been actively engaged in designing various polycyclic cage compounds. These molecules display unusual structural properties that differ from normal limits. Consequently, our understanding of their design, chemical behaviour, and reactivities are greatly enhanced and the synthesis and reactions of strained molecules occupy a special place in organic chemistry. Polycyclic cage molecules are now routinely used as useful building blocks and are also produced in multigram quantities using a variety of contemporary synthetic methods. With these advances, cage systems have become useful synthons in designing natural and non-natural targets.[1] [2]

The oxiranes derived from acetophenones are useful precursors to design conazoles.[3] The Corey–Chaykovsky epoxidation (CCE) was used to produce intricate oxirane intermediates. The epoxidation reaction was conducted with trimethylsulfoxonium iodide in an aqueous NaOH solution and under microwave (MW) irradiation conditions (Scheme [1]).[4a] To produce 1-(1H-indol-1-yl)-2-phenyl-3-(1H-1,2,4-triazol-1-yl)-propan-2-ols 3ad and 4ae as antifungal agents, oxiranes 1 and 2 were opened regioselectively with a various indole derivative (Scheme [1]).

Zoom Image
Scheme 1 Synthesis of antifungal agent via Corey–Chaykovsky reaction. Reagents and conditions: (a) K2CO3,1H-1,2,4-triazole, CH3CN, MW 85 °C (50 W), 50 min. (b) NaOH (aq), TMSOI, toluene, MW 80 °C (50 W), 50 min. (c) NaH, indole derivative, DMSO, r.t., 12 h.

Epoxides are known to undergo an interesting transformation to produce carbonyl compounds, diols and amino alcohols under acidic or basic conditions. In view of this we are interested in preparing various epoxides containing cage compounds.[4b]

Simmons and Park prepared macrocyclic diammonium ions in 1968.[5] [6] [7] Here, they introduced the term in/out stereoisomerism and called the process of conformational interconversion of these isomers homeomorphic isomerism.[8] Later, Lehn reported the first cryptands which utilize the inside lone pairs on bridgehead nitrogen atoms in the complexation of metal ions,[9] [10] [11] [12] [13] [14] [15] [16] and the interest in such type of compounds grow strongly ever since. In/out isomerism in polycyclic cage compounds occurs when the relative positions of the atoms within the cage structure are different. These atoms can either be oriented towards the inside of the cage (in) or towards the outside of the cage (out). This aspect involving different spatial arrangements of the atoms within the molecule, leading to distinct isomeric forms (Figure [1]).

Zoom Image
Figure 1 Pictorial representation of in/in, out/out, and in/out isomers

McMurry and co-workers[17] used the route to in/out bicyclic olefin for the synthesis of 6 from the ketoaldehyde 5 (Scheme [2]). They also performed the catalytic hydrogenation to generate 7 and later they generalized this method to produce several of these bridgehead olefins which are useful as precursors for the novel carbocations with bent three-center, two-electron C–H–C bonds.[18] Although rare, this is certainly an interesting structural and stereochemical phenomenon. In/out isomerism has also been observed in a number of natural products.[19]

Zoom Image
Scheme 2 Hydrocarbon with in/out C–H bonds

Recently, substituted pentacyclo[5.4.0.02,6.03,10.05,9]undecanes (PCUDs) have drawn the attention of synthetic chemists in the production of cyclopentanoid natural products.[20] [21] [22] To expand the chemical space of pentacyclo[5.4.0.02,6.03,10.05,9]undecanes,[23] [24] [25] [26] [27] [28] we choose Cookson’s dione[29f] 8 to synthesize new cage diepoxides which can exist in three different stereochemical dispositions (Figure [2]). The synthesis of cage diepoxides derived from Cookson’s dione 8 represents a new class of building blocks, and the availability of such synthons opens up new possibilities for further synthetic manipulations. The strategic use of the CCE reaction allows the introduction of epoxide functionalities at specific positions within the cage molecular framework. This reaction involves the use of sulfur ylides, generated in situ from sulfonium or sulfoxonium salts, which undergo nucleophilic attack on the carbonyl carbon of Cookson’s dione 8, leading to the formation of a diepoxide product. The dione 8 containing two keto groups in a congested environment provide a unique opportunity for the controlled generation of in/in, in/out, and out/out diepoxides. The distinctive feature of this synthetic approach lies in the ability to access three isomeric forms of the cage diepoxide, each exhibiting unique structural disposition and hopefully different reactivity pattern.

Zoom Image
Figure 2 In/in, in/out, and out/out isomers of cage diepoxides derived from Cookson’s dione 8

These cage diepoxides can serve as useful precursors for various synthetic transformations such as ring-expansion methodologies. Some of these polycyclic cage compounds have inherent ring strain and are also symmetrical in nature.[29] Their molecular architecture has drawn our attention due to the lack of conformational mobility. The CCE sequence has been used for the preparation of diepoxide and its isomers are useful building blocks in designing functionalized polycyclic cage molecules starting with Cookson’s dione such as 8.

We identified a new synthetic approach to inner- and outer-cage epoxides using the CCE and the Peterson olefination reaction as key steps.[1] In this regard, Cookson’s dione 8, easily prepared from the inexpensive reagents such as p-benzoquinone and cyclopentadiene in two steps, is used as the starting material and it was treated with sulfonium ylide or sulfoxonium ylide which gave the ‘in’ monoepoxide 12 and ‘in/in’ diepoxide 9 (Scheme [3]).[30]

Zoom Image
Scheme 3 Formation of monoepoxide and diepoxide

To obtain the diepoxide 9 as major product, we have tried different reaction conditions. In this regard, we observed the best results when using dry DMSO and NaH conditions in 92% and 90% yields, respectively. After various conditions (Table [1]), entries 4 and 8 gave the best results.

Table 1 Synthesis of Monoepoxide and Diepoxide from Cookson’s Dione 8

Entry

Reagent

Ylide (equiv)

Temp (℃)

Solvent

Time (h)

Yield of 12 (%)a

Yield of 9 (%)a

1

Me3S(=O)+I, NaH

3

25

THF

5

NRb

NRb

2

Me3S(=O)+I, NaH

3

reflux

THF

12

NRb

NRb

3

Me3S(=O)+I, NaH

3

25

DMSO

5

23

62

4

Me3S(=O)+I, NaH

5

25

DMSO

3

0c

92

5

Me3S+I, NaH

3

25

THF

6

NRb

NRb

6

Me3S+I, NaH

3

reflux

THF

12

NRb

NRb

7

Me3S+I, NaH

3

25

DMSO

4

25

65

8

Me3S+I, NaH

5

25

DMSO

3

0c

90

9

Me3S(=O)+I, NaH

5

25

DMSO–THF

15

NRb

NRb

10

Me3S+I, NaH

5

25

DMSO–THF

15

NRb

NRb

a Isolated yield.

b No reaction.

c Yield of compound 12.

We have successfully synthesized novel derivatives of epoxides from Cookson’s dione 8. By employing the CCE sequence later, we then generalized this new route with other cage diones.

For example, cage compounds 13, 15, and 17, prepared along similar lines, were subjected to the CCE sequence and gave the corresponding epoxides 14, 16 and 18, respectively. 13C NMR spectra of these epoxides 9 indicated the presence of symmetry. It is interesting to note that in the case of cage compounds containing spirocyclopentane and spirocyclopropane units (compounds 14, 16, and 18) all CH2 units are not equivalent due to the tetrahedral nature of the spiro center, and cyclopentane and cyclopropane moieties are placed orthogonally at the spiro center. This strategic approach can open up new avenues in the development of advanced cage compounds (Table [2]).

Table 2 List of Various Diepoxide Derivatives Prepared via Corey–Chaykovosky Reaction

Entry

Carbonyl compd

Sulfooxonium ylide

Diepoxide product

Conditionsa

Time (h)

Yield (%)

1

8

9

NaH, DMSO
(0 ℃ to rt)

3

92

2

13

14

NaH, DMSO
(0 ℃ to rt), 4 h

4

80

3

15

16

NaH, DMSO
(0 ℃ to rt)

3

74

4

17

18

NaH, DMSO
(0 ℃ to rt)

3

78

a Base was added at 0 ℃ then reaction was allowed to stir at 25 ℃.

Since we aimed to prepare all the possible isomers of epoxides, we changed our strategy and prepared the other two isomers (in/out and out/out). To synthesize the in/out and out/out diepoxides, we performed the selective olefination of the dicarbonyl cage compound 8 to generate the known keto-olefin product 22 using the Peterson olefination conditions. A convenient modification of the Peterson reaction was introduced by Chan and Chang,[31] who reacted magnesium alkoxides of β-hydroxy silanes with acetyl chloride (or thionyl chloride) in situ to form alkenes directly. In the case of the reaction with acetyl chloride, the proposed intermediate 3-acetoxy silane was not isolated.[30] Later, Marchand and co-workers reported the formation of monoalkene with (trimethylsilyl)methylmagnesium chloride in dry THF to get the monoolefinated product.

To synthesize the keto-olefin 22, we started our journey from the Cookson’s dione 8 as starting material. When Cookson’s dione 8 was treated with 1 equivalent of (trimethylsilyl)methylmagnesium chloride, and the resulting product was quenched with acetyl chloride, compound 21 was produced in good yield instead of the expected keto-olefin 22 (Scheme [4]).

Zoom Image
Scheme 4 Selective conversion of diketone into monoketone by Peterson olefination

This outcome suggests that the intermediate magnesium alkoxide 19 undergoes an intramolecular nucleophilic addition with the remaining carbonyl group, leading to the formation of a hexacyclic intermediate magnesium alkoxide 20. Later, subsequent trapping of 20 by the addition of acetyl chloride results in the formation of the compound 21.[1b] Compound 21 could be converted into keto-olefin 22 cleanly by two pathways. Refluxing 21 in THF with trifluoroacetic acid (1.2 equiv) for 60 h yielded 22 in good yield. Alternatively, a more convenient method involves stirring 21 in methylene chloride with boron trifluoride etherate at room temperature for 24 h. An intriguing aspect of this reaction sequence is the selective conversion of one of the two identical carbonyl groups present in symmetrical dione 8 into an unsymmetrical keto-olefin system 22.

Zoom Image
Scheme 5 Epoxidation by m-CPBA

By subjecting the keto-olefin 22 to m-CPBA oxidation or hydrogen peroxide treatment, we noticed the formation of side products such as 24 or 25, prompting the need for a modified approach (Scheme [5]). To address this issue, we initially performed the CCE of the keto-olefin 22, resulting in the formation of monoepoxide 26. Subsequent treatment of the epoxide 26 with m-CPBA yielded diepoxide as in/out configuration. This strategy involving a sequential epoxidation allowed the selective synthesis of both monoepoxide 26 and in/out diepoxide isomer 10, minimizing the formation of undesired byproducts such as 24 and 25 (Scheme [6]).

Zoom Image
Scheme 6 Synthesis of the in/out diepoxide

Having prepared the in/out epoxide, the keto-olefin 22 was reacted with (trimethylsilyl)methylmagnesium chloride, resulting in the formation of the diolefin 27. Subsequent treatment of 27 with m-CPBA led to the production of the diepoxide predominantly, the out/out isomer indicating a specific stereochemical arrangement of the substituents around the newly formed epoxide rings (Scheme [7]).[1]

Zoom Image
Scheme 7 Formation of the out/out diepoxide

The CCE products have been found to be useful substrates for the Demjanov rearrangement (DR) sequence, allowing for the generation of ring-expanded products. The reaction involves the rearrangement of epoxide, resulting in the ring-expansion sequence (Scheme [8]).[32]

Zoom Image
Scheme 8 Demjanov rearrangement

Later, we proceeded to treat the diepoxide with lithium bromide (LiBr) and lithium iodide (LiI) in an attempt to expand the five-membered ring of compound 10. However, despite conducting these reactions individually, we encountered difficulties in isolating and characterizing the products. The yield obtained was low, which posed challenges in identifying and analyzing the formed products. We need to explore new conditions to improve the yield and facilitate the isolation and characterization of products (Scheme [9]).[33]

Zoom Image
Scheme 9 Attempt towards ring homologation

When in/in diepoxide 9 was treated with liquid NH3 for the ring-opening reaction, the starting material was recovered (Scheme [10]).[32]

Zoom Image
Scheme 10 Attempts to open the epoxide ring

Similarly, the out/out isomer also did not open as indicated by our initial experimentation (Scheme [11]).

Zoom Image
Scheme 11 Attempts to open the epoxide ring

In view of the above considerations, we conceived a new approach to produce the dione 31 a useful precursor to hexaprismane 32 by ring-homologation sequence of Cookson’s dione 8 (Scheme [12]).[34] We are committed to further expand our efforts in this direction.

Zoom Image
Scheme 12 Retrosynthetic path to 31

We have also performed DFT calculations using the B3LYP[35] functional and 6-31G**[36] basis set in the Gaussian suite of programs to compute the stability, optimized structures, and the HOMO–LUMO gap of 9, 10, and 11. The DFT-optimized structures are well reproduced with the X-ray structure (Table [3] and Table [4]).

Table 3 Few Important Bond Length Parameters from the Single-Crystal X-ray Structure

Bond length

C1–C2 (Å)

C1–O1 (Å)

C2–O1 (Å)

C3–C4 (Å)

C3–O2 (Å)

C4–O2 (Å)

9

1.463

1.435

1.459

1.463

1.435

1.459

10

11

1.458

1.445

1.450

1.458

1.450

1.451

Table 4 Few Important Bond Length Parameters from the DFT-Optimized Structure

Bond length

C1–C2 (Å)

C1–O1 (Å)

C2–O1 (Å)

C3–C4 (Å)

C3–O2 (Å)

C4–O2 (Å)

9

1.469

1.425

1.437

1.469

1.425

1.437

10

1.468

1.431

1.439

1.461

1.437

1.442

11

1.465

1.436

1.438

1.465

1.437

1.438

All computed IR frequencies are positive and no imaginary frequencies are found which suggests the correct optimization with global minimum (see the Supporting Information, Tables S3–S5). The computed electronic energy suggests that compound 10 is more stable than 11, and 9 lies at 0.726 kJ mol–1 and is 12.172 kJ mol–1 higher in energy (see Table [5]).

Table 5 Relative Energy of Compounds 9, 10, and 11

Cage diepoxides

Relative energy (kJ mol–1)

10

0.000

11

0.726

9

12.172

We have also plotted the (HOMO–1) HOMO and LUMO (LUMO+1) energy gap and it suggests that the HOMO of compound 11 is lower in energy than the other two while it has a LUMO with very high energy. Hence the HOMO–LUMO gap of 11 is the highest. The HOMO–LUMO gap follows the following order 11 > 10 > 9. While the HOMO of each compound consists of a C–C σ-bond and lone pair orbital of oxygen atoms, the LUMO is mainly composed of a C–C σ* orbital (see Figure [3]).

Zoom Image
Figure 3 Comparison of frontier molecular orbitals (FMO) of compounds 9, 10, and 11

In summary, the successful synthesis of the three diepoxides 9, 10, and 11 derived from cage compounds related to Cookson’s dione using the CCE sequence and m-CPBA oxidation demonstrate the utility of our strategy. In view of the diverse applications of dione 8 in organic synthesis, the methodology developed here may find useful applications in the synthesis of new strained molecules. This strategy allows the controlled introduction of epoxide moieties, providing access to intricate molecular architectures. The versatility extends beyond structural diversity, offering opportunities for ring expansion. The accessibility of these epoxides opens up new possibilities for new cage skeletons.


#

Conflict of Interest

The authors declare no conflict of interest.

Supporting Information


Corresponding Author

Sambasivarao Kotha
Department of Chemistry
Indian Institute of Technology-Bombay, Powai, 400076
India   

Publication History

Received: 02 July 2024

Accepted after revision: 12 August 2024

Accepted Manuscript online:
12 August 2024

Article published online:
10 September 2024

© 2024. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany


Zoom Image
Scheme 1 Synthesis of antifungal agent via Corey–Chaykovsky reaction. Reagents and conditions: (a) K2CO3,1H-1,2,4-triazole, CH3CN, MW 85 °C (50 W), 50 min. (b) NaOH (aq), TMSOI, toluene, MW 80 °C (50 W), 50 min. (c) NaH, indole derivative, DMSO, r.t., 12 h.
Zoom Image
Figure 1 Pictorial representation of in/in, out/out, and in/out isomers
Zoom Image
Scheme 2 Hydrocarbon with in/out C–H bonds
Zoom Image
Figure 2 In/in, in/out, and out/out isomers of cage diepoxides derived from Cookson’s dione 8
Zoom Image
Scheme 3 Formation of monoepoxide and diepoxide
Zoom Image
Scheme 4 Selective conversion of diketone into monoketone by Peterson olefination
Zoom Image
Scheme 5 Epoxidation by m-CPBA
Zoom Image
Scheme 6 Synthesis of the in/out diepoxide
Zoom Image
Scheme 7 Formation of the out/out diepoxide
Zoom Image
Scheme 8 Demjanov rearrangement
Zoom Image
Scheme 9 Attempt towards ring homologation
Zoom Image
Scheme 10 Attempts to open the epoxide ring
Zoom Image
Scheme 11 Attempts to open the epoxide ring
Zoom Image
Scheme 12 Retrosynthetic path to 31
Zoom Image
Figure 3 Comparison of frontier molecular orbitals (FMO) of compounds 9, 10, and 11