Synthesis of [2.2]Paracyclophane‐Based Glycidic Amides Using Chiral Ammonium Ylides

An ammonium ylide-mediated stereoselective protocol for the synthesis of a series of novel [2.2]paracyclophanebased epoxides starting from racemic 4-formyl[2.2]paracyclophane has been developed. By using achiral ammonium salts as ylide precursors, the corresponding epoxide products were obtained in isolated yields up to 76% and with diastereoselectivities up to d.r.=9:1. When carrying out the reaction with chiral ammonium salts instead, the products were accessible with e.r.>93.5 :6.5 and d.r.>3:1, accompanied with a moderately enantioselective kinetic resolution of the racemic starting aldehyde (e.r.=75:25).


Results and Discussion
We started our investigations by reacting the achiral trimethylammonium salts 4 with rac-2 under liquid/ liquid or liquid/solid biphasic conditions ( Table 1 gives an overview of the most significant results obtained hereby; trimethylamine was used as an amine leaving group as it was found to be the group of choice for aldehydes 3 recently [34] ). First experiments using amide 4a (Y = NEt 2 ) with Cs 2 CO 3 as a solid base in different solvents (entries 1-3) showed that the targeted epoxide 6a is indeed accessible by such a strategy, but in low yields only. While toluene did not allow for any product formation (entry 1), i PrOH favored the Cannizzaro disproportionation of aldehyde 2 and reactions in CH 2 Cl 2 were found to be rather slow with this carbonate base. Changing for liquid/liquid conditions with an excess of a stronger hydroxide base, [33,34] yield of 6a could be increased significantly. With respect to the relative configuration of product 6a, we observed formation of two trans-diastereoisomers, differing in the configuration of the planar chiral paracyclophane unit. Later investigations with chiral ammonium salts (vide infra, Scheme 2) suggest that the major diastereomer has a (2S,3R,S p )-rel-configuration, as depicted in the general structure of compounds 6 shown in Table 1. Although KOH and NaOH both gave approximately the same yield and d.r. after one day at 25°C (entries 4 and 5), reactions with KOH gave larger amounts of unidentified side products, while the NaOH-mediated reaction was in general slower in conversion, but also cleaner. Gratifyingly, by prolonging the reaction time to three days and increasing the temperature to 40°C, the yield could be improved significantly (72 %) without reducing the diastereoselectivity (d.r. = 9 : 1; entry 6).
With these conditions at hand, we next investigated the use of different ammonium salts 4. Upon testing different amides (entries 6 -12), it turned out that secondary amides (like compound 4f, entry 11) or the Weinreb amide 4g (entry 12) are not suited, which is in line with the limitations observed previously when using simple arylaldehydes for such reactions. [33] In contrast, tertiary amides like compounds 4b-4d performed reasonably well (entries 7-9). Interestingly, the dimethylamide-based 4e was found to be rather insoluble and decomposed quickly under the standard conditions and in this case, it was only possible to access traces of the product 6e by using t BuOK in DMSO instead (entry 10). Noteworthy however, this class of amide later on performed much better when using a chiral ammonium salt auxiliary (vide infra, Scheme 2). Unfortunately, ester-(entry 13) or ketonebased (entry 14) ammonium salts 4h and 4i did not allow for any product formation at all, which again comes as no surprise upon comparison with our own previous observations for arylaldehydes 3, where these ylide-precursors were found to be not reactive as they form more stable and thus less reactive ammonium ylides. [32] Having identified suitable conditions for the racemic synthesis of different tert. amide-containing epoxides 6 with reasonable yields and good diastereoselectivities, we next examined whether we could render this reaction enantioselective by using chiral aminebased ammonium salts 4. It should be emphasized that the identification of a suited chiral amine that allows for satisfying yields and enantioselectivities for such ammonium ylide-mediated epoxidations was a major challenge previously, [32] mainly because of the significantly lower leaving group quality of chiral tert. amines like Cinchona alkaloids compared to the simple achiral amines (i. e. trimethylamine). [32] In our inves-tigations of the epoxidation of a variety of benzaldehyde derivatives 3 the chiral bicyclic proline-derived amines A were the only systems that allowed for product 5-formation with good yields and selectivities, while Cinchona alkaloids totally failed. [32] Thus it also came as no surprise that only ammonium salts 4 A containing the chiral auxiliaries A allowed for any formation of epoxide 6a herein (Scheme 2,A; Cinchona alkaloids were tested too but did not allow for any product formation). Amongst the four tested amine leaving groups A1 -A4, the cyclohexyl-containing derivative A4 was found to be the best suited one. Using one equivalent of the A4-containing ammonium salt 4a A4 at room temperature gave highly enantioenriched 6a (e.r. > 96.5 : 3.5 for both diastereomers; higher temperature led to lower selectivities) accompanied with a good d.r. of 5 : 1 after one day (30 % conversion of rac-2). In order to reach a 'KR-suited' conversion of roughly 50 %, the reaction was run for two days next (resulting in 60 % conv.), giving 6a with a slightly reduced d.r. of 3 : 1 as well as a slightly lower e.r. (i. e. for the major diastereomer). In addition, this approach allowed for the recovery of (R p )-2 (assignment of the absolute configuration was carried out by comparison of the measured (À )-rotation with previous reports [17] ) with moderate enantioenrichment (e.r. = 75 : 25). With these results in hand, the configuration of the major stereoisomer of epoxide 6a formed through this strategy was assigned as follows: Considering the predominant (R p )-configuration of the recovered aldehyde 2, the [2.2]paracyclophane unit within product 6a has to be (S p )-configurated. In addition, our recent studies revealed that ammonium salts 4 A4 strongly favor the formation of (2S,3R)epoxides when reacted with arylaldehydes 3. Considering the observed high levels of enantioselectivity controlled by the auxiliary A4 for both, reactions with aldehyde 2 and aldehydes 3, it can therefore be assumed that the major stereoisomer of product 6a isolated herein has (2S,3R,S p )-configuration, as shown in Scheme 2. Based on the only moderate enantioenrichment of recovered (R p )-2, it can be assumed that the KR is the less-selective process herein (compared to epoxide-formation) and therefore the minor diastereomer of epoxide 6a should have (2S,3R,R p )-configuration instead, which also explains why the d.r. decreases with increasing conversion of rac-2.
With this reasonably enantio-and diastereoselective protocol in hand, we finally tested the use of different amide-based ammonium salts 4 A4 (Scheme 2,B). In contrast to the racemic protocol (and as already stated before), the chiral dimethylamide-based salt 4e A4 performed well in this asymmetric approach, giving product 6e with high selectivity and in good isolated yield. In addition, also the piperidine-based epoxide 6b could be obtained in a similarly efficient manner.

Conclusions
An ammonium ylide-mediated protocol for the formation of novel [2.2]paracyclophane-based epoxides 6 starting from the racemic aldehyde 2 has been developed. When using achiral ammonium salts 4 as ylide precursors, the corresponding amide-containing epoxides 6 were obtained in isolated yields up to 76 % with high diastereoselectivities up to d.r. = 9 : 1. When carrying out the reaction with the chiral ammonium salts 4 A instead, the epoxides 6 were accessible with high enantio-and diastereoselectivities, accompanied with a moderately enantioselective kinetic resolution of the racemic starting aldehyde 2.

Experimental Section
General Experimental Details 1 H-and 13 C-NMR spectra were recorded on a Bruker Avance III 300 MHz spectrometer with a broad band observe probe and a sample changer for 16 samples which is property of the Austro Czech NMR Research Center 'RERI uasb'. NMR Spectra were referenced on the solvent peak and chemical shifts are given in ppm. High resolution mass spectra were obtained using a Thermo Fisher Scientific LTQ Orbitrap XL with an Ion Max API Source. Analyses were made in the positive ionization mode if not otherwise stated. Purine (exact mass for [M + H] + = 121.050873) and 1,2,3,4,5,6-hexakis(2,2,3,3-tetrafluoropropoxy)-1,3,5,2,4,6-triazatriphosphinane (exact mass for [M + H] + = 922.009798) were used for internal mass calibration. HPLC was performed using a Thermo Scientific Dionex Ultimate 3000 system with diode array detector with a CHIRAL ART Cellulose-SB (250 × 4.6 mm, 5 μm) chiral stationary phase. Optical rotations were recorded on a Schmidt + Haensch Polarimeter Model UniPol L1000 at 589 nm. All chemicals were purchased from commercial suppliers and used without further purification unless otherwise Scheme 2. Asymmetric epoxidation of aldehyde 2 using chiral ammonium salts 4.

General Procedure for the Racemic Epoxidation
The achiral ammonium salt 4 (0.1 mmol, 1 equiv.) was dissolved in CH 2 Cl 2 (1 mL, 0.1 M) followed by the addition of NaOH (50 % aq., 550 μL, 100 equiv.) and rac-2 (25 mg; 1 equiv.). The mixture was heated to 40°C and stirred for 3 d. After cooling to room temperature, the mixture was extracted with CH 2 Cl 2 and the combined organic phases dried over Na 2 SO 4 . The crude product was purified by column chromatography (elution of the recovered aldehyde with heptanes/AcOEt 10 : 1 and of the epoxides with heptanes/ AcOEt 2 : 1).

Supplementary Material
Supporting information (copies of NMR spectra and HPLC traces) for this article is available on the WWW under https://doi.org/10.1002/hlca.202100073.