Asymmetric Synthesis of 1,2-Limonene Epoxides by Jacobsen Epoxidation

Keywords

Jacobsen epoxidation - 1,2 limonene epoxides - asymmetric synthesis

Introduction

In recent years, cannabis and some of its bioactive components have received increasing attention in basic research and pharmaceutical applications.[1] Among these, cannabinoids including cannabidiol (CBD), tetrahydrocannabinol (THC), cannabichromene, and cannabigerol have shown extensive pharmacological effects. As early as the 1980s, Dronabinol (marketed as Marinol) was launched to prevent chemotherapy-induced nausea and vomiting, which was a synthetic form of delta-9-tetrahydrocannabinol (Δ⁹-THC). In 2018, Epidiolex (CBD oral solution) was approved by the Food and Drug Administration to be the first CBD-based product available on the U.S. market for the treatment of two rare forms of epilepsy—Lennox–Gastaut syndrome and Dravet syndrome—which are among the two most difficult types of epilepsy to treat.[2] To date, THC and CBD have been the most studied cannabinoids.

The main way to obtain cannabinoids is to separate them from the dry substances and fresh cannabis leaves. Synthesizing cannabinoids by chemical synthesis instead of natural extraction has also become a research hotspot. In the reported total synthesis routes of CBD or THC, p-mentha-2,8-dien-1-ol (5) was used as an intermediate.[3] Among the listed cannabinoid drugs, compound 5 is also the key part of their structure ([Scheme 1]). Compound 5 is synthesized by four steps of epoxidation, ring opening, oxidation, and elimination from compound 1 ([Scheme 2]).[4] In the reported methods, compound 2 is a diastereomeric mixture ([Scheme 3]), which is difficult to obtain a single configuration by fractionation or column chromatography purification. In the ring-opening reaction of compound 2, the trans-epoxide was selectively opened with aqueous dimethylamine to generate 3. The cis-2 remained largely unreacted to affect the purity of the compound 3, and so it is difficult to obtain compound 5 with high optical purity.[4] Therefore, it is necessary to explore the asymmetric oxidation synthesis of compound 1 to obtain trans-1,2-limonene epoxide with high optical purity for the synthesis of target cannabinoid drugs.

Scheme 1 p-Mentha-2,8-dien-1-ol as a common building block in bioactive cannabinoids.

Scheme 2 Synthesis of p-mentha-2,8-dien-1-ol (5). Reagents and conditions: (i) m-CPBA, CHCl₃, 0°C, 2 h, 62%; (ii) 40% HNMe₂ (aq), 80°C, 18 h, 88% based on the trans-isomer; (iii) 30% H₂O₂, 50% CH₃CN (aq), r.t., 2 h, 100%; (iv) 180°C, 1 mm Hg, 74%.

Scheme 3 The structures of cis-2 and trans-2.

Jacobsen epoxidation is an asymmetric epoxidation of olefins without specified functional groups. The chiral salen–metal complexes are used as Jacobsen's enanitioselective epoxidation catalysts. The commonly used oxidants are iodosyl benzene (for organic solvents) and sodium hypochlorite (for water media).[5] In addition, hydrogen peroxide and m-CPBA can also be used as oxidants for this reaction, simultaneously additional ligands are required, such as 4-methylmorpholine N-oxide (NMO).[6] Despite the widespread application and the utility of the Jacobsen method, the optimum reaction conditions for its enantioselectivity have remained obscure.

To our knowledge, Montes de Correa and colleagues have engaged the challenge of asymmetric epoxidation of (R)-(+)-limonene with the salen–manganese complex as a catalyst to obtain 1,2-limonene epoxides by applying Jacobsen's epoxidation method. They found that the product stereochemistry was strongly dependent on the absolute configuration of both the catalyst and the limonene. The combination of R-(+)-limonene with (R,R)-Jacobsen catalyst or (S)-(−)-limonene with (S,S)-Jacobsen catalyst formed a matched pair, giving rise to diastereomeric excess values of 56 and 45%, respectively.[7] Ratnasamy and colleagues have reported that Mn (salen) complexes immobilized on sulfonic acid-functionalized SBA-15 exhibited efficient catalytic activity for selective epoxidation of R-(+)-limonene with aerial oxygen. 1,2-Limonene epoxide was the major product. However, the diastereomeric excess for the endo-enantiomer was only 39.8%.[8] Bernardo-Gusma and colleagues have reported asymmetric epoxidation of R-(+)-limonene (1) using the Jacobsen catalysts in organic solvents and ionic liquids. R-(+)-Limonene (1) was selectively converted to 1,2-epoxi-p-ment-8-enes with a diastereoselectivity of 70% in organic solvents and 74% in ionic liquids.[9] Asymmetric epoxidation of limonene has been reported in many studies, but no high diastereomeric excess of 1,2-epoxides has been obtained.

In this article, R-(+)-limonene (1) was used as the substrate to screen the chiral salen-metal catalysts, oxidants, axial ligands, and dosage of ligands for asymmetric oxidation reactions. And we have successfully selected suitable conditions to prepare cis- and trans-epoxides with high optical purity, which can be used as important raw materials for the preparation of related cannabinoid drugs. The results provide a useful reference for the total synthesis of cannabinoids.

Methods and Experiments

Methods for (+)-1,2-Limonene Oxide Quantification

All reaction medium samples were diluted in methanol and analyzed in a gas chromatography–mass spectrometry (GC-MS; Agilent 7890B GC-5977A), equipped with a HP-5 column (30 m length × 0.25 mm internal diameter × 0.25 µm film thickness), and a mass (MS) detector (Agilent 5977A MSD). The samples were injected into the column initially at 50°C; after a holding time of 2 minutes, the temperature was increased to 15°C/min until 250°C, with a final holding time of 5 minutes.

Experiment for the Synthesis of 2 (a Diastereomeric Mixture)

A solution of m-CPBA (16 mmol) in DCM (30 mL) was added dropwise to a solution of R-(+)–limonene (1) (10 mmol) in DCM (30 mL) over 30 minutes in such a way that the temperature did not rise over 5°C. The solution was then stirred at 0°C for 30 minutes, and then at room temperature for 1 hour before the addition of sodium hydroxide (1 mol/L, 20 mL, 20 mmol). The organic phase was collected, washed with sodium carbonate and brine, and then dried over anhydrous magnesium sulfate. The solvent was removed by rotary evaporation and the residue was purified by column chromatography (petroleum ether:ethyl acetate = 50:1) to obtain 1,2-epoxide (2) (yield: 62%) as a colorless oil. [α]²⁰ _D: +38.3 (0.1, CHCl₃); GC-MS (m/z): 152.1 (M⁺); ¹H NMR (600 MHz, CDCl₃) δ 4.73–4.66 (m, 4H), 3.05 (s, 1H), 2.99 (d, J = 5.4 Hz, 1H), 2.15–2.07 (m, 2H), 2.05–2.01 (m, 2H), 1.91–1.78 (m, 4H), 1.71 (dd, J = 11.3, 3.8 Hz, 2H), 1.69 (s, 3H), 1.66 (s, 3H), 1.63 (s, 1H), 1.53 (dddd, J = 10.1, 5.4, 3.7, 2.0 Hz, 1H), 1.39–1.35 (m, 2H), 1.31 (s, 3H), 1.30 (s, 3H).

Typical Procedure for Jacobsen Asymmetric Oxidation Reaction

To a solution of R-(+)-limonene (1) (10 mmol), Jacobsen's catalyst (0.5 mmol), and axial ligand (30 mmol) in 30 mL DCM was added the m-CPBA (16 mmol, in 30 mL of DCM) drop by drop, and the resulting mixture was vigorously stirred at 0°C for 10 hours. After completion of the reaction, the mixture was detected by GC-MS. Saturated sodium bicarbonate solution was added to the reaction solution. The DCM layer was collected, washed with water, and dried over anhydrous sodium sulfate. The residue was purified by column chromatography to obtain 1,2-epoxide as a colorless oil.

Experiment for the Synthesis of cis-2

To a solution of R-(+)-limonene (1) (10 mmol), catalyst 7 (0.5 mmol), and NMO (50 mmol) in 30 mL DCM was added m-CPBA (16 mmol, in 30 mL of DCM) drop by drop, and the mixture was vigorously stirred at 0°C for 10 hours. After completion of the reaction, the mixture was detected by GC-MS. Saturated sodium bicarbonate solution was added to the reaction solution. The DCM layer was collected, washed with water, and dried over anhydrous sodium sulfate. The residue was purified by column chromatography (petroleum ether:ethyl acetate = 50:1) to obtain cis-2 (yield: 48.2%) as a colorless oil. [α]²⁰ _D: +70.7 (0.1, CHCl₃); GC-MS (m/z): 152.1 (M⁺); ¹H NMR (400 MHz, CDCl₃) δ 4.69 (dd, J = 13.0, 11.5 Hz, 2H), 3.02 (d, J = 2.2 Hz, 1H), 2.16–1.97 (m, 2H), 1.89–1.76 (m, 2H), 1.69 (dd, J = 7.5, 5.1 Hz, 1H), 1.67 (s, 3H), 1.66–1.61 (m, 1H), 1.57–1.47 (m, 2H), 1.28 (s, 3H); ¹³C NMR (101 MHz, CDCl₃) δ 148.51 (s), 108.57 (s), 60.05 (s), 56.85 (s), 35.73 (s), 30.25 (s), 28.15 (s), 25.44 (s), 23.81 (s), 20.62 (s).

Experiment for the Synthesis of trans-2

To a solution of R-(+)-limonene (1) (10 mmol), catalyst 6 (0.5 mmol), and 2-pyridinol-1-oxide (30 mmol) in 30 mL DCM was added m-CPBA (16 mmol, in 30 mL of DCM) drop by drop, and the mixture was vigorously stirred at 0°C for 10 hours. After completion of the reaction, the mixture was detected by GC-MS. Saturated sodium bicarbonate solution was added to the reaction solution. The DCM layer was collected, washed with water, and dried over anhydrous sodium sulfate. The residue was purified by column chromatography (petroleum ether:ethyl acetate = 60:1) to obtain trans-2 (yield: 36.3%) as a colorless oil. [α]²⁰ _D: +79.1 (0.1, CHCl₃); GC-MS (m/z): 152.2 (M⁺); ¹H NMR (600 MHz, CDCl₃) δ 4.65 (s, 2H), 2.97 (d, J = 5.4 Hz, 1H), 2.01 (ddd, J = 15.0, 7.2, 4.3 Hz, 2H), 1.86 (ddd, J = 14.8, 12.0, 6.1 Hz, 1H), 1.72–1.65 (m, 5H), 1.36 (ddd, J = 12.2, 8.2, 3.8 Hz, 2H), 1.30 (s, 3H); ¹³C NMR (151 MHz, CDCl₃) δ 149.21 (s), 109.08 (s), 59.27 (s), 57.50 (s), 40.74 (s), 30.75 (s), 29.88 (s), 24.34 (s), 23.09 (s), 20.22 (s).

Results and Discussion

R-(+)-Limonene (1) was used as the substrate, and first, the oxidants and catalysts used in Jacobsen asymmetric epoxidation ([Scheme 4]) were screened (the results are shown in [Table 1]). Initially, when 1 reacted with 3 equiv. of H₂O₂, epoxidation did not occur with or without the axial ligand NMO and catalyst ([Table 1], entries 1 and 2). By replacing H₂O₂ with m-CPBA as oxidants, the reaction could be performed, and a mixture of isomers with cis- to trans- ratios close to 1:1 was obtained ([Table 1], entry 3, [Scheme 5]). When m-CPBA was used as the oxidant, 7, 8, 9 as the catalyst, and NMO as the axial ligand, excess cis-isomer epoxides ([Table 1], entries 5–7) were obtained, and the highest diastereoselectivity was found with 7 as the catalyst, and the diastereomeric excess was up to 98% ([Table 1], entry 5, [Scheme 6]). And when 6 was used as the catalyst, excess trans-isomer epoxides were obtained with diastereomeric excess of 52% ([Table 1], entry 4).

Scheme 4 The catalysts used in the Jacobsen method.

Scheme 5 GC-MS chromatogram showing the direct epoxidation by m-CPBA. GC-MS, gas chromatography–mass spectrometry.

Scheme 6 GC-MS chromatogram showing epoxide products starting from R-(+)-limonene (1) using 7 as the catalyst. GC-MS, gas chromatography–mass spectrometry.

Table 1 Screening of catalysts and oxidants[a]

Entry	Catalyst	Oxidant	Axial ligands	de %[b]
1	–	H2O2 [c]	–	–
2	6	H2O2 [c]	NMO	–
3	–	m-CPBA	–	7[d]
4	6	m-CPBA	NMO	52[e]
5	7	m-CPBA	NMO	98[d]
6	8	m-CPBA	NMO	41[d]
7	9	m-CPBA	NMO	37[d]

Abbreviation: de, diastereomeric excess.

^a All the reactions were performed at 0°C in DCM (60 mL) with alkene (10 mmol), NMO (50 mmol, if necessary), catalysts (0.5 mmol, 5.0 mmol%) and oxidants (16 mmol), unless otherwise.

^b Determined by GC-MS. The order of peaks of cis-2 and trans-2 referred to Mccue et al[14] and Melchiors et al[15].

^c 3 equiv. of H₂O₂ was used.

^d Referred to cis-1,2-limonene oxide (predominant epoxide).

^e Referred to trans-1,2-limonene oxide (predominant epoxide).

The disparate results in asymmetric induction can be understood in terms of the common model proposed by Jacobsen epoxidation. Olefins attack from the side of the metal–oxygen bond in the Jacobsen asymmetric oxidation reaction.[10] [11] When metal atoms are complexed with axial ligands, they are closer to the salen plane, and the interaction between olefins and substituents on salen ligands is stronger. The complexation of axial ligands can also reduce the reactivity of oxygenated salen complexes to improve the selectivity.[12]

According to the above epoxidation mechanism analysis, to obtain the trans-2 with higher diastereomeric excess, we screened the amount of NMO ([Table 2]). However, GC-MS showed that the diastereomeric excess value of trans-epoxide was not significantly increased by increasing the amount of NMO. When the ligand dosage was 3 equiv., the diastereomeric excess value was only 53% ([Table 2], entry 5), indicating that NMO was not the best ligand for the oxidation system.

Table 2 Screening of NMO dosage[a]

Entry	NMO (equiv.)	de %[b]
1	0	18[c]
2	0.5	0
3	1	24[d]
4	2	49[d]
5	3	53[d]
6	5	52[d]
7	10	48[d]

Abbreviation: de, diastereomeric excess.

^a All the reactions were performed at 0°C in DCM (60 mL) with alkene (10 mmol), catalyst 6 (0.5 mmol, 5.0 mmol%), and m-CPBA (16 mmol).

^b Determined by GC-MS. The order of peaks of cis-2 and trans-2 referred to Mccue et al[14] and Melchiors et al[15].

^c Referred to cis-1,2-limonene oxide (predominant epoxide).

^d Referred to trans-1,2-limonene oxide (predominant epoxide).

Considering the importance of axial ligands in the epoxidation systems, to obtain trans-2 with higher diastereomeric excess value, we further performed a series of screening of axial ligands reported in the epoxidation systems ([Table 3]).[13] It was found that high purity trans-epoxides with a diastereomeric excess of 94% could be obtained successfully when 2-hydroxypyridine-N-oxide (HOPO) was used as the axial ligand ([Table 3], entry 5; [Scheme 7]).

Scheme 7 GC-MS chromatogram showing epoxide products starting from R-(+)-limonene (1) using 6 as the catalyst, 2-pyridinol-1-oxide (HOPO) as the axial ligand. GC-MS, gas chromatography–mass spectrometry.

Table 3 Screening of axial ligands[a]

Entry	Axial ligands	de %[b]
1	NMO	53
2	Imidazole	20
3	2-Methylimidazole	1
4	1-Methylimidazole	23
5	2-Hydroxypyridine-N-oxide (HOPO)	94
6	Piperidine	8
7	N-Methyl piperazine	10
8	Pyridine-1-oxide	35
9	4-tert-Butylpyridine	15

Abbreviation: de, diastereomeric excess.

^a All the reactions were performed at 0°C in DCM (60 mL) with alkene (10 mmol), catalyst 6 (0.5 mmol, 5.0 mmol%), m-CPBA (16 mmol), and axial ligands (30 mmol).

^b Determined by GC-MS. The order of peaks of cis-2 and trans-2 referred to Mccue et al[14] and Melchiors et al[15]; Referred to trans-1,2-limonene oxide (predominant epoxide).

Conclusion

In summary, we have successfully found an effective method of catalytic asymmetric epoxidation to synthesize cis-1,2-limonene epoxide and trans-1,2-limonene with high diastereomeric excess values, respectively. cis-1,2-Limonene epoxide with a diastereomeric excess of 98% was synthesized by asymmetric oxidation of 7 as the catalyst, NMO as the axial ligand and m-CPBA as the oxidant. Using 6 as the catalyst, 2-pyridinol-N-oxide as the axial ligand, and m-CPBA as the oxidant, the trans-2 could be obtained with a diastereomeric excess of 94%. In this study, we reported for the first time that 1,2-limonene epoxides in rather high diastereoselectivity (>90%) were obtained by Jacobsen epoxidation, which will provide an effective preparation method of key intermediates for the chemical synthesis of cannabinoid drugs.

• References

• 1 Nelson KM, Bisson J, Singh G. et al. The essential medicinal chemistry of cannabidiol (CBD). J Med Chem 2020; 63 (21) 12137-12155
• 2 Elsaid S, Kloiber S, Le Foll B. Effects of cannabidiol (CBD) in neuropsychiatric disorders: a review of pre-clinical and clinical findings. Prog Mol Biol Transl Sci 2019; 167: 25-75
• 3 Pirrung MC. Synthetic access to cannabidiol and analogs as active pharmaceutical ingredients. J Med Chem 2020; 63 (21) 12131-12136
• 4 Wilkinson SM, Price J, Kassiou M. Improved accessibility to the desoxy analogues of Δ9-tetrahydrocannabinol and cannabidiol. Tetrahedron Lett 2013; 54 (01) 52-54
• 5 Shi QP, Shi ZH, Li NG. et al. Asymmetric epoxidation of olefins with homogeneous chiral (salen) manganese (III) complex. Curr Org Chem 2013; 17 (23) 2936-2970
• 6 Ballistreri FP, Gangemi CM, Pappalardo A, Tomaselli GA, Toscano RM, Trusso Sfrazzetto G. (Salen)Mn(III) catalyzed asymmetric epoxidation reactions by hydrogen peroxide in water: a green protocol. Int J Mol Sci 2016; 17 (07) 1112
• 7 Cubillos J, Vargas M, Reyes J, Villa A, Montes De Correa C. Effect of the substrate and catalyst chirality on the diastereoselective epoxidation of limonene using Jacobsen-type catalysts. Chirality 2010; 22 (04) 403-410
• 8 Saikia L, Srinivas D, Ratnasamy P. Chemo-, regio- and stereo-selective aerial oxidation of limonene to the endo-1,2-epoxide over Mn(Salen)-sulfonated SBA-15. Appl Catal A Gen 2006; 309 (01) 144-154
• 9 Pinto LD, Dupont J, de Souza RF, Bernardo-Gusma K. Catalytic asymmetric epoxidation of limonene using manganese Schiff-base complexes immobilized in ionic liquids. Catal Commun 2008; 9 (01) 135-139
• 10 Brandes BD, Jacobsen EN. Highly enantioselective, catalytic epoxidation of trisubstituted olefins. J Org Chem 1994; 59 (16) 4378-4380
• 11 Kürti L, Blewett MM, Corey EJ. Origin of enantioselectivity in the Jacobsen epoxidation of olefins. Org Lett 2009; 11 (20) 4592-4595
• 12 Pietikäinen P. Catalytic and asymmetric epoxidation of unfunctionalized alkenes with hydrogen peroxide and (Salen)Mn(III) complexes. Tetrahedron Lett 1994; 35 (06) 941-944
• 13 Bagherzadeh M, Jonaghani MA, Amini M. et al. Synthesis of dipyroromethanes in water and investigation of electronic and steric effects in efficiency of olefin epoxidation by sodium periodate catalyzed by manganese tetraaryl and trans disubstituted porphyrin complexes. J Porphyr Phthalocyanines 2019; 23 (06) 671-678
• 14 Mccue AJ, Wells R, Anderson JA. Confirmation of chirality in homogeneous and heterogeneous salen-based catalysts. ChemCatChem 2011; 3 (04) 699-703
• 15 Melchiors M, Vieira T, Pereira LP. et al. Epoxidation of R-(+)-limonene to 1,2-limonene oxide mediated by low-cost immobilized Candida antarctica lipase fraction B. Ind Eng Chem Res 2019; 58 (31) 13918-13925