Challenges in the Highly Selective [3 + 1]-Cycloaddition of an Enoldiazoacetamide to Form a Donor–Acceptor Cis-Cyclobutenecarboxamide †

A substituted donor–acceptor cyclobutenecarboxamide is synthesized with modest enantiocontrol through a chiral copper(I) complex catalyzed [3 + 1]-cycloaddition reaction of α-acyl diphenylsulfur ylides with 3-siloxy-2-diazo-3-butenamides. With a methyl substituent on the 4-position of the 3-butenamide, the cis-vicinal-3,4-disubstituted cyclobutenecarboxamide is formed with >20:1 diastereocontrol. Donor-acceptor 3-methyl-2-siloxycyclopropenecarboxamide is rapidly formed from the reactant enoldiazoamide and undergoes catalytic ring opening to give only the Z-γ-substituted metallo-enolcarbene. Elimination from 3-siloxy-2-diazo-3-pentenamide to form the conjugated 3-siloxy-2,4-pentadienamide is competitive but minimized at low temperature.

Prior research has suggested that diazoamides are more stable and more selective in their catalytic reactions emanating from metal carbene intermediates [19,20], and we anticipated that this selectivity could be applied to catalytic [3 + 1]-cycloaddition of silyl group protected enoldiazoacetamides. However, initial efforts indicated that the same conditions and catalysts that were effective with enoldiazoacetates were not as productive or selective with enoldiazoacetamides. In particular, reactions with α-acyl dimethylsulfur Molecules 2021, 26, 3520 2 of 13 ylides gave low yields for the cycloaddition product, and stereoselectivities were low. To enhance both the efficiency of the transformation and its selectivity to produce substituted donor-acceptor cyclobutenecarboxamides with exceptional stereocontrol, we undertook a comprehensive effort to optimize reactants, conditions, and catalyst ligands to achieve high yields, as well as high enantioselectivities and diastereocontrol (Figure 1c).
Molecules 2021, 26, x 2 of 13 or selective with enoldiazoacetamides. In particular, reactions with α-acyl dimethylsulfur ylides gave low yields for the cycloaddition product, and stereoselectivities were low. To enhance both the efficiency of the transformation and its selectivity to produce substituted donor-acceptor cyclobutenecarboxamides with exceptional stereocontrol, we undertook a comprehensive effort to optimize reactants, conditions, and catalyst ligands to achieve high yields, as well as high enantioselectivities and diastereocontrol (Figure 1c).

Results and Discussion
We began our investigation with the cycloaddition of TIPS-protected N,N-dimethylenoldiazoacetamide 1a with α-benzoyl dimethylsulfur ylide 2a using the same copper(I) catalyst and chiral ligand (L1, Scheme 1) that were most effective in reactions with enoldiazoacetates ( Figure 2) [18]. However, reaction at room temperature under the same conditions produced the [3 + 1]-cycloaddition product 3a in 58% yield having 0% ee after complete dinitrogen extrusion of 1a (Table 1, entry 1). Since phenyl in place of methyl increases the reactivity of the sulfur ylide [21], α-benzoyl diphenylsulfur ylide 2b was prepared [22][23][24][25], and its reaction with 1a under the same conditions gave 3a in 85% yield with 17% ee (Table S1, entry 2, Supplementary Materials). The major enantiomer was assigned to be R based on its correlation in sign of rotation and relative retention volume by HPLC compared with the [3 + 1]-cycloaddition product with the corresponding enoldiazoacetate [18].

Results and Discussion
We began our investigation with the cycloaddition of TIPS-protected N,N-dimethylenoldiazoacetamide 1a with α-benzoyl dimethylsulfur ylide 2a using the same copper(I) catalyst and chiral ligand (L1, Scheme 1) that were most effective in reactions with enoldiazoacetates ( Figure 2) [18]. However, reaction at room temperature under the same conditions produced the [3 + 1]-cycloaddition product 3a in 58% yield having 0% ee after complete dinitrogen extrusion of 1a (Table 1, entry 1). Since phenyl in place of methyl increases the reactivity of the sulfur ylide [21], α-benzoyl diphenylsulfur ylide 2b was prepared [22][23][24][25], and its reaction with 1a under the same conditions gave 3a in 85% yield with 17% ee (Table S1, entry 2, Supplementary Materials). The major enantiomer was assigned to be R based on its correlation in sign of rotation and relative retention volume by HPLC compared with the [3 + 1]-cycloaddition product with the corresponding enoldiazoacetate [18]. ules 2021, 26,

Ligand Control of Enantioselectivity in the [3 + 1]-Cycloaddition of 2-diazo-N,N-dimethyl-3-(triisopropylsiloxy)but-3-enamide
Using the 4-phenyl-Sabox L1, the ester analog of 1a, methyl enoldiazoacetate, was able to reach 83% ee in its [3 + 1]-cycloaddition reaction with 2b. That 17% ee (Table S1, entry 2) could be achieved in reactions with N,N-dimethyl-enoldiazoacetamide 1a was surprising and not easily explained as due to the size of dimethylamido relative to methoxy groups, nor by the expected electronic influence of amide relative to ester groups on the diazo carbon. Consequently, we directed our attention to expanding our search for ligands that might increase enantioselectivity with a survey of Box (bis-oxazoline) and SaBox (sidearmed bis-oxazoline) ligands (Scheme 1).

Ligand Control of Enantioselectivity in the [3 + 1]-Cycloaddition of 2-diazo-N,N-dimethyl-3-(triisopropylsiloxy)but-3-enamide
Using the 4-phenyl-Sabox L1, the ester analog of 1a, methyl enoldiazoacetate, was able to reach 83% ee in its [3 + 1]-cycloaddition reaction with 2b. That 17% ee (Table S1, entry 2) could be achieved in reactions with N,N-dimethyl-enoldiazoacetamide 1a was surprising and not easily explained as due to the size of dimethylamido relative to methoxy groups, nor by the expected electronic influence of amide relative to ester groups on the diazo carbon. Consequently, we directed our attention to expanding our search for ligands that might increase enantioselectivity with a survey of Box (bis-oxazoline) and SaBox (sidearmed bis-oxazoline) ligands (Scheme 1).

Ligand Control of Enantioselectivity in the [3 + 1]-Cycloaddition of 2-Diazo-N,N-dimethyl-3-methyl-2-(triisopropylsiloxy)-1-cyclopropenecarboxamide
The discovery that there was such a substantial reverse in diastereoselectivity in reactions performed with the enoldiazoacetamide (1b) from that found with the corresponding enoldiazoacetate was surprising and prompted investigation of the effect of the ligand L on diastereocontrol. To avoid perturbations in selectivities arising from the different reactivities and selectivities of E-and Z-1b, we chose to use donor-acceptor cyclopropene 5 as the carbene source. Previous results have shown that the donor-acceptor cyclopropene undergoes ring opening with transition metal catalysts to form only the Z-metallo-enolcarbene isomer [31]. We surveyed a representative series of Box and Sabox ligands for [3 + 1]cycloaddition with 2b, and these results are given in Table 2. As previously stated, yields of [3 + 1]-cycloaddition products are limited by the competing formation of diene 4. However, even with this limitation, L14 (Table 2, entry 6) produced Z-3b with a 12:1 Z:E ratio at room temperature and allowed the isolation of pure Z-3b in 67% yield with 71% ee, and a higher yield, dr and % ee were achieved when the reaction was performed at −20 • C (Table 2, entry 7).

General Information
All reactions, unless noted, were performed in oven-dried (120 • C) glassware with magnetic stirring under an inert atmosphere of dry nitrogen. Analytical thin layer chromatography was carried out using EM Science silica gel 60 F254 plates (MilliporeSigma, Burlington, MA, USA); visualization was accomplished with UV light (254 nm). Column chromatography was performed on CombiFlash ® Rf200 and Rf+ purification systems (Teledyne Technologies, Thousand Oaks, CA, USA) using normal phase disposable columns. 1 H-NMR spectra were recorded on a Bruker spectrometer (500 MHz, Bruker Corporation, Billerica, MA, USA). Chemical shifts were reported in ppm downfield from tetramethylsilane with the solvent resonance as the internal standard (CDCl 3 , δ = 7.28). Spectra were reported as follows: chemical shift (δ ppm), multiplicity (br = broad singlet, s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, comp = composite of magnetically non-equivalent protons, dd = doublet of doublets, td = triplet of doublets; dt = doublet of triplets), coupling constants (Hz), integration and assignment. 13 C-NMR spectra were collected on Bruker instrument (125 MHz, Bruker Corporation) with complete proton decoupling. Chemical shifts are reported in ppm from the tetramethylsilane with the solvent resonance as internal standard (CDCl 3 , δ = 77.0). High-resolution mass spectra (HRMS) were performed on a Bruker MicroTOF-ESI mass spectrometer (Bruker Corporation) with an ESI resource using CsI or LTQ ESI positive ion calibration solution as the standard. Enantioselectivities were determined by HPLC analysis at 25 • C using an Agilent 1260 Infinity HPLC System (Agilent-Technologies, Santa Clara, CA, USA) equipped with an G1311B quaternary pump, G1315D diode array detector, G1329B auto-sampler, G1316A thermostated column compartment and G1170A valve drive. For instrument control and data processing, Agilent OpenLAB CDS ChemStation Edition (1200-series) for LC & LC/MS Systems (Rev. C.01.07 [26]) software was used. Chiralpak OD-H (0.46 mm × 250 mm) columns were obtained from Daicel Chiral Technologies (Chiral Technologies Inc., West Chester, PA, USA). Tetrahydrofuran, dichloromethane, chloroform, and toluene were purified using a JC Meyer solvent purification system. All other solvents were purified and dried using standard methods.
purification system. All other solvents were purified and dried using standard methods.

General Procedure for Asymmetric Catalytic [3 + 1]-Cycloaddition to Prepare Z-3b and E-3b from 5
To an oven-dried sealable 2-dram vial equipped with a stir bar were added CuOTf·Tol 1/2 (2.6 mg, 0.010 mmol, 5 mol%) and bisoxazoline ligand L14 (6.1 mg, 0.012 mmol, 6 mol%). After the vial was evacuated and backfilled with N 2 three times, dry DCM (1.0 mL) was added via a syringe and the resulting solution was stirred at room temperature for 1 h before sulfur ylide 2b (61 mg, 0.20 mmol, 1.0 equiv.) dissolved in dry DCM (0.5 mL) was added dropwise via a syringe for 1 h. The reaction was stirred at room temperature for 5 min and then a solution of cyclopropenecarboxamide 5 (71 mg, 0.24 mmol, 1.2 equiv.) in dry DCM (0.5 mL) was then added dropwise. The reaction mixture was stirred at the same temperature for 24 h, filtered through a short pad of silica gel and washed with hexanes/EtOAc (1:1, 10 mL). The filtrate was further concentrated and directly subjected to analysis by 1 H-NMR. After that, the residue was purified by flash chromatography on silica gel using 10:0 to 10:3 hexanes:ethyl acetate as the eluent to afford to afford the expected [3 + 1]-cycloaddition product Z-3b, E-3b and diene (4).
The competitive formation of diene 4 was another surprise in this transformation. Not previously observed from reactions with the corresponding enoldiazoacetates [18], this product was observed in variable amounts from all reactions catalyzed by copper(I) with Box or Sabox ligands. However, diene 4 (Scheme S2, Supplementary Materials) was not formed when the copper catalyzed reaction with ylide 2b was performed without ligand with either enoldiazoacetamide Z-1b or with donor-acceptor cyclopropenecarboxamide 5, and neither the sulfur ylide nor the ligand alone caused diene formation in reactions performed over 24 h. Since a 20% excess of ligand was used to ensure that unligated copper(I) was minimized, we thought that the diene might arise from the ligand, acting as a base, to effect elimination from the intermediate metallo-enol carbene. Indeed, using 2.4 mol% triethylamine in place of the ligand resulted in diene formation (29% yield, Scheme S3, Supplementary Materials). However, limiting the amount of excess ligand in the [3 + 1]-cycloaddition reaction to exactly 1:1 correspondence with the copper catalyst did not reduce the amount of diene 4 formation; but lowering the temperature to −20 • C brought diene formation down to 2%.
In summary, Cu(I) catalyzed asymmetric [3 + 1] cycloaddition of α-benzoyl diphenylsulfur ylide 2b with 3-methyl-or un-substituted cyclopropenecarboxamides gives access to the synthesis of donor-acceptor cyclobutenecarboxamides in good yield and moderate enantioselectivity. Unlike their corresponding enoldiazoacetate, the γ-methyl substituted amide gives a high preference for the cis diastereoisomer. Reactivity and stereoselectivity of the amide and ester are significantly different, and formation of diene 4 from amide 1b suggests an elimination pathway for the intermediate metallo-enolcarbene.