Diastereoselective synthesis of chiral 1,3-cyclohexadienals

A novel approach to the production of chiral 1,3-cyclohexadienals has been developed. The organocatalysed asymmetric reaction of different β-disubstituted-α,β-unsaturated aldehydes with a chiral α,β-unsaturated aldehyde in the presence of a Jørgensen-Hayashi organocatalyst provides easy and stereocontrolled access to the cyclohexadienal backbone. This method allows for the synthesis of potential photoprotective chiral 1,3-cyclohexadienals and extra extended conjugation compounds in a simple manner.


Introduction
Organocatalysis is one of the fastest growing areas in organic chemistry [1][2][3][4]. The enantioselective organocatalytic Diels-Alder reaction from the seminal communication of Prof. Mac-Millan et al. [5] constitutes one of the most interesting research areas. The synthesis of enantiomerically enriched building blocks is an important task in organic synthesis, where cyclohexadienes [6][7][8][9][10][11] are of special interest due to their reactivity. Although the use of monosubstituted α,β-unsaturated aldehydes is more extended, in the last few years the use of β-disubstituted-α,β-unsaturated aldehydes has become more prevalent in this area. There are numerous examples of asymmetric synthesis by using organocatalysis, as shown by the work of Professor Serebryakov et al. in the synthesis of cyclohexa-1,3-dienes from prenal and unsaturated esters or derivatives, [12][13][14][15][16] Professor Hong et al. for the synthesis of aromatic aldehydes by organocatalytic [4+2] or [3+3] cycloaddition of α,β-unsaturated aldehydes [17][18][19] and Professor Watanabe et al. in citral, 1, dimerization. [20][21][22][23][24][25] The cyclohexadienal scaffold has been shown to be bioactive in numerous cases throughout the literature. For example, the citral dimer shows antibiotic activity [26] and the retinal dimer could contribute to macular degeneration. [27] As chiral aldehyde 2 has been intensively used as a synthetic building block in the synthesis of bioactive natural products, [28][29][30][31][32][33] this study sought to obtain chiral cyclohexadienals using 2 in combination with different β-methyl disubstituted-α,β-unsaturated aldehydes in the presence of different catalysts (5)(6)(7)(8)(9)(10), which avoid the dimerization of these compounds (Fig 1). a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 In the last few decades the potentially dangerous effects of UV radiation exposure have been extensively demonstrated [34][35][36]. While UVC light is filtered by the upper atmospheric layers, UVB and UVA light penetrate the upper layers of the atmosphere and reach the Earth's surface. Photoprotection against this radiation can prevent skin damage and deleterious effects on DNA. However, it is important not to overdo protection against UVB as this can reduce the biosynthesis of vitamin D [37,38]. Therefore, photoprotective agents that selectively absorb UVB and UVA radiation are the UV-filters needed for developing effective and safe sunscreens.
There are two groups of UV filters: inorganic and organic compounds. The inorganic filters scatter, reflect or absorb UV radiation, however, only TiO 2 and ZnO are FDA approved. The organic UV filters consist of structurally simple aromatic molecules that absorb in UVA and UVB. The organic UV filters used in sunscreens, and approved by the FDA (Fig 2)[39] can be classified as cinnamates, benzophenones, PABA and salicilate derivatives and others. Despite their use in sunscreens, there are several studies regarding the toxicity, and especially the phototoxicity, of these compounds [40][41][42][43][44][45][46].
In this work, cyclohexadienals containing different substitutions have been synthesized as easily accessible high-conjugated compounds with interesting UV-Vis properties, making them suitable for use as photoprotective UV-filters.

Materials and methods
All reactions were performed in oven-dried glassware under positive Ar pressure with magnetic stirring, unless otherwise noted. Air and moisture-sensitive liquids and solutions were transferred via a syringe or a stainless-steel cannula. TLC was performed on 0.25 mm E. Merck silica gel 60 F254 plates and visualized under UV light (λ = 254 nm) or by staining with potassium permanganate. Flash chromatography was performed on E. Merck 230-400 mesh silica gel 60. All reagents were purchased from commercial suppliers, and used without further purification unless otherwise noted. Solvents were distilled from suitable drying agents (CaH 2 or Na wire) under an Ar atmosphere at 760 mmHg. All moisture-and/or oxygen-sensitive solids were handled and stored in a glove box under N 2 . The NMR spectra were recorded on Bruker AVANCE 400 MHz DRX and Varian Mercury 200 MHz using CDCl 3 as solvent. NMR data is reported as follows: chemical shift (δ) (parts per million, ppm); multiplicity: s (singlet), d (doublet), t (triplet), q (quartet) and br (broad); coupling constants (J) are given in Hertz (Hz). 1 H NMR chemical shifts were calibrated with respect to residual chloroform in CDCl 3 centered at 7.26 ppm, whereas for 13 C NMR, the center peak for CDCl 3 , centered at 77.0 ppm, was used for the calibration. The IR spectra were obtained on a Shimadzu IR Affinity-1 (film over NaCl). All NMR and IR spectra can be found in S1 File. The HRMS spectra were obtained on an Applied Biosystems QSTAR XL mass spectrometer. The optical rotation was performed on a Perkin-Elmer 241 digital polarimeter using cuvette with l = 1 dm and CHCl 3 as the solvent. Absorbance measures were determined in 200-700 nm region using iPrOH as the solvent and an UV quartz cuvette (l = 1 cm) in a Shimadzu UV-2401PC spectrophotometer with thermostatic system at 20˚C. The UV-Vis spectra can be found in S4 File.
The different experimental conditions of the catalyst, solvent and additives tested are shown in Table 1.
When using a non-chiral organocatalyst, such as pyrrolidine, 5, cyclohexadienal 4 was obtained in low yields, although without diastereoselectivity (entry 1). The use of L-proline, 6 (entries 2-3), using different solvents, or no solvent at all, gave the required cyclohexadienal 4 in very low yields and the citral dimer 3, as a subproduct. Then, MacMillan´s organocatalysts 7 and 8 were tested, but no result was obtained (entries [4][5]. In addition, the Jørgensen-Hayashi catalysts 9 and 10 were used in different solvent conditions, obtaining different results depending on the solvent used, ranging from moderate yields of cyclohexadienal 4 (entries 6-9 and 22) to no reaction at all (entries [10][11][12][13][14]. As can be seen in Table 1, in some cases the reaction was carried out in presence of additives such as acids (BzOH, o-nitro-BzOH, AcOH, TsOH, (±)-1,19-binaphthyl-2,29-diyl hydrogenphosphate[(±) BINAP-OH] or TFA) and bases (DBU) (entries 15-21) with improved yields. The best result was obtained when the Jørgensen-Hayashi catalyst 9 in CHCl 3 as the solvent, was used without any additional additive (entry 8) and produced a moderate yield and a good d.r. C-6 in compound 12 was established as S, because of NOE between H1' and H6 did not appear. Later on, the absolute configuration was confirmed by X-Ray of an analogue (24a).

Synthesis of chiral cyclohexadienals with different substituents
The mechanism could be understood by a Diels-Alder reaction, as suggested by Serebryakov et al. [12][13][14][15][16] and Watanabe et al. [20]. Similarly, this will would explain that the stereochemistry Hexane -4a,b (5) 80:20 Toluene -4a,b (20) 75:25 (19) 60:40 (27) 80:20 1,3-cyclohexadienals obtained in the final product does not depend on the Z or E stereochemistry of the α,β-unsaturation of the aldehyde used in the reaction. The same result was obtained with E-citral or a mixture E/Z-citral. E-citral was obtained from geraniol as described in the literature. [48] Once the conditions for the synthesis of cyclohexadienals were obtained, the generality of the reaction using different β-disubstituted-α,β-unsaturated aldehydes and 2 as starting materials was then observed, Fig 5 and Table 2.
The reaction was initiated using a simple α,β-unsaturated aldehyde such as 13. When catalysts 9 or 10 were used, both produced a good yield and diastereoselection. When catalyst 10 was used, instead of 9, the yield slightly decreased but diastereoselection remained complete. When using aromatic aldehydes, the reaction worked very well, especially with the p-methoxyphenyl group (entries 5-6) which produced excellent yields and diastereoselection with both catalysts 9 and 10. When a bromophenyl group was used (entries 7-10), the yield and diastereoselection decreased but when a p-nitrophenyl group (entries [11][12] was used the yield increased with both catalysts and the diastereoselection was excellent, especially with catalyst 10. When the reaction was run using an aliphatic cyclic aldehyde, such as catalyst 19, the yield was very poor (entry 13) but diastereoselection was complete. As can be seen in Table 2, the reaction proceeded quite well, especially when using aromatic aldehydes.

Crystallographic analysis of cyclohexadienal 24a
Compound 24a was crystallized. In Fig 6, the X-ray crystal structure of compound 24a [49] is shown and confirms the stereochemistry of compound 24a at C-6. The stereochemistry of this compound was previously predicted by the NMR of compound 12, and by analogy, the stereochemistry of compounds 20 to 26 was established.

UV-Vis absorption analysis
The UV-Vis absorbance of different photostable cyclohexadienals was measured (Table 3 and S4 File) in order to test the possible application of these compounds as photoprotective agents.
The majority of the compounds at concentrations in the order of 10 −6 absorbed UVA and UVB. Compound 21b exhibited values suitable for photoprotection against UVA owing to the higher area under the curve (AUC) at that particular wavelength region and its molar extinction coefficient (ε = 13200 M -1 cm -1 ). The best results found in the UVB region were shown by compound 23b which had an extinction coefficient of 34700 M -1 cm -1 at 288nm. However, the compound that was able to better absorb UVA and UVB was 23a, with molar extinction coefficients of 8000 M -1 cm -1 in UVA and 10900 M -1 cm -1 in UVB.
A global view of UV absorption of this chiral aromatic cyclohexadienal can be seen in

Synthesis
General procedure for the optimization of conditions for cyclohexadienals (4a,b). Catalyst 5-10 (0.5 eq) were added to a solution containing 2 (0.3 mmol, 1 equiv.) and 1 (0.3 1,3-cyclohexadienals mmol, 1 equiv.) in solvent (1.5 mL, 0.2M) at r.t. The reaction mixture was stirred at r.t. for 48h. The solution was concentrated in and the residue was purified by flash column chromatography (EtAcO:hexane) to obtain cyclohexadienals 4a and 4b as a yellow oil and dimer 3 as a colourless oil.   Catalyst 9 (0.5 eq) was added to a solution containing 2 (0.15 mmol, 1 equiv.) and E-citral (0.15 mmol, 1 equiv.) in CHCl 3 (0.75 mL, 0.2M) at r.t. The reaction mixture was stirred at r.t. for 48h. The solution was concentrated in vacuum and the residue was purified by flash column chromatography (EtAcO:hexane) to obtain a mixture of cyclohexadienals 4a and 4b as a yellow oil (yield 37%; d.r. 85:15).
p-TsOH (21 mg, 0.11 mmol) was added to a solution containing 11 (35mg, 0.11 mmol) and MeOH (1.5 mL). The reaction mixture was stirred at r.t. for 14h. The reaction was quenched with H 2 O. The crude mixture was extracted with EtOAc (3x10 mL). The combined organic layers were washed with H 2 O, sat. NaHCO 3 solution and brine, dried over Na 2 SO 4 , filtered and concentrated under vacuum to yield 12 (