Convenient Synthesis of Pyrazolo[4′,3′:5,6]pyrano[4,3-c][1,2]oxazoles via Intramolecular Nitrile Oxide Cycloaddition

A simple and efficient synthetic route to the novel 3a,4-dihydro-3H,7H- and 4H,7H-pyrazolo[4′,3′:5,6]pyrano[4,3-c][1,2]oxazole ring systems from 3-(prop-2-en-1-yloxy)- or 3-(prop-2-yn-1-yloxy)-1H-pyrazole-4-carbaldehyde oximes has been developed by employing the intramolecular nitrile oxide cycloaddition (INOC) reaction as the key step. The configuration of intermediate aldoximes was unambiguously determined using NOESY experimental data and comparison of the magnitudes of 1JCH coupling constants of the iminyl moiety, which were greater by approximately 13 Hz for the predominant syn isomer. The structures of the obtained heterocyclic products were confirmed by detailed 1H, 13C and 15N NMR spectroscopic experiments and HRMS measurements.


Introduction
The 1,3-dipolar cycloaddition reaction of nitrile oxides as 1,3-dipoles and alkenes/ alkynes as dipolarophiles has become an efficient tool in organic synthesis to obtain various substituted isoxazolines/isoxazoles [1][2][3]. The reaction was developed by Rolf Huisgen and described by Albert Padwa in their investigations on 1,3-dipolar cycloadditions [4,5]. Nitrile oxides, which are typically generated in situ, undergo subsequent 1,3-dipolar cycloaddition to form appropriate isoxazoles or isoxazolines. Numerous methods of nitrile oxide generation have been reported, mainly including the dehydration of nitroalkanes [6][7][8] and oxidation of aldoximes [9][10][11]. Alternatively, Svejstrupor described the synthesis of isoxazolines and isoxazoles from hydroxyimino acids via the visible-light-mediated generation of nitrile oxides by two sequential oxidative single electron transfer processes [12]. More recently, Chen et al. reported the synthesis of fully substituted isoxazoles from nitrile oxides, which were generated in situ from copper carbene and tert-butyl nitrite [13].
Notably, the intramolecular nitrile oxide cycloaddition (INOC) reaction can provide a route for the preparation of isoxazoles or isoxazolines annulated to various carbo-or heterocycles. For example, the intramolecular 1,3-dipolar cycloaddition of 2-phenoxybenzonitrile N-oxides to neighboring benzene rings, accompanied by dearomatization, formed the corresponding isoxazolines in high yields [14]. Recently, a method for the stereoselective synthesis of novel isoxazoline/isoxazole-fused indolizidine-, pyrrolizidine-and quinolizidinebased iminosugars has been developed, employing N-alkenyl/alkynyl iminosugar Cnitromethyl glycosides as nitrile oxide precursors in 1,3-dipolar cycloaddition reactions [15]. The phthalate-tethered INOC strategy has also been described as a novel method for the synthesis of 12-15-membered chiral macrocycles having a bridged isoxazoline moiety in a highly regio-and diastereoselective manner [16]. Furthermore, diversity-oriented access isoxazoline moiety in a highly regio-and diastereoselective manner [16]. Furthermore, diversity-oriented access to isoxazolino and isoxazolo benzazepines as possible bromodomain and extra-terminal motif protein (BET) inhibitors has been reported via a post-Ugi heteroannulation involving the intramolecular 1,3-dipolar cycloaddition reaction of nitrile oxides with alkenes and alkynes [17]. In addition, an intramolecular 1,3-dipolar nitrile oxide cycloaddition strategy has been applied as an efficient synthesis protocol for the regio-and diastereoselective construction of highly functionalized tricyclic tetrahydroisoxazoloquinolines [18].
Over the years, the isomerism of aldoximes has been thoroughly studied and many different NMR-based approaches have been developed, mainly due to the large differences in chemical shifts, coupling constants and distinct through-space connectivities in NOESY measurements of aldoxime syn-anti isomers [37]. The configurational assignment of aldoximes 4a-c was relatively easy due to the presence of both isomers, as it is well established that the resonance of the iminyl-H proton in the syn isomer is greatly shifted upfield by approximately δ 0.5-0.7 ppm in the 1 H NMR spectra compared to the anti isomer [38]. Moreover, the 1D selective NOESY experimental data of aldoximes 4a-c showed that, upon irradiation of the hydroxyl proton N-OH of the predominant syn isomer, a strong positive NOE on the pyrazole 5-H proton was observed, while the minor isomer showed a positive NOE on the iminyl-H, therefore confirming the anti configuration. Finally, a heteronuclear 2D J-resolved NMR experiment was used in order to determine 1 J CH coupling constants throughout the series of aldoximes. It is well established from previous studies that there is a large and constant difference between the magnitudes of 1 J CH coupling constants of the iminyl moiety in syn-anti isomers [39], which is larger by at least 10-15 Hz for the syn isomer. The measurements of compounds 4a-c showed that the relevant 1 J CH coupling constants of the iminyl moiety were around 175.0 Hz for the predominant syn isomer, while the minor anti isomer provided significantly lower coupling constant values by around 13.0 Hz. The configuration of aldoxime 4d as a pure syn isomer was easily deduced from NOESY measurements and the 1 J CH coupling constant of the iminyl moiety, which was 174.5 Hz. The analysis of 15 N NMR spectroscopic data showed highly consistent chemical shift values within each isomer, in a range from δ −18.2 to −25.7 ppm in the case of the syn isomer and in a range from δ −15.6 to −16.5 ppm for the anti isomer. A comparison of the relevant NMR data of aldoximes is presented in Table S1.
The formation of 3a,4-dihydro-3H,7H-and 4H,7H-pyrazolo[4′,3′:5,6]pyrano[4,3c] [1,2]oxazole ring systems was easily deduced after an in-depth analysis of NMR spectral data, which were obtained through a combination of standard and advanced NMR spectroscopy techniques, such as 1 H- 13    In the case of compound 5a, the multiplicity-edited 1 H-13 C HSQC spectrum allowed us to identify the pairs of geminally coupled methylene protons, since both protons displayed cross-peaks with the same carbon. For instance, it showed two pairs of negative signals at δ H 4.66, 3.79 and 4.78, 4.17 ppm, which have one-bond connectivities with the methylene carbons C-3 (δ 69.7 ppm) and C-4 (δ 70.9 ppm), respectively. The chemical shifts of these methylene groups are expected to be similar and downfield compared to a neighboring methine group at site 3a, because both are bound to the oxygen atoms O-2 and O-5. This adjacent protonated carbon C-3a (δ 46.7 ppm) relative to the aforementioned methylene sites was easily assigned from an appropriate correlation in the 1 H-13 C H2BC spectrum.
In the 1 H-15 N HMBC spectrum of 5a, strong long-range correlations between the methylene 3-H proton at δ 4.66 ppm and the 3a-H proton at δ 3.86-3.91 ppm with the oxazole N-1 nitrogen at δ −32.2 ppm were observed. The lack of long-range correlations with another pair of methylene protons (δ 4.78, 4.17 ppm), and the aforementioned N-1 nitrogen, strongly hinted at assigning this methylene group to site 4. In order to unambiguously discriminate between these methylene groups, the 1 H-13 C heteronuclear couplings were measured using a 1 H-13 C J-HMBC experiment, thus providing complimentary evidence for correct structural assignment. The J-HMBC spectrum showed a strong correlation between the methylene proton δ 4.78 ppm and the quaternary carbon C-5a with an 8.0hertz coupling constant, while the proton δ 4.66 ppm correlated very weakly, with a J value of only 2.2 Hz, which was attributed to a 5 J C-5a, H-3 . Finally, the pyrazole 8-H proton (δ 8.13 ppm) not only exhibited long-range HMBC correlations with neighboring N-7 "pyrrole-like" (δ −177.4 ppm) and N-6 "pyridine-like" (δ −117.7 ppm) nitrogen atoms, but also with the C-5a, C-8a and C-8b quaternary carbons, which were unambiguously assigned with the subsequent 1,1-ADEQUATE experiment, thus allowing all the heterocyclic moieties to be connected together. The structure of compounds 5b-d was determined by analogous NMR spectroscopy experiments, as described above. The skeleton of the pyrazolo[4 ,3 :5,6]pyrano[4,3-c][1,2]oxazole ring system contains three nitrogen atoms. The chemical shifts of the N-1, N-6 and N-7 atoms of compounds 5a-c were in a range from δ −30.9 to −32.2, δ −116.9 to −117.7 and δ −177.4 to −179.5 ppm, respectively, while in the case of compound 5d, which lacked a phenyl moiety at site 7, the chemical shifts of N-1, N-6 and N-7 atoms were δ −35.8, δ −112.3 and δ −194.4 ppm, respectively.
In the case of compound 6, a comparison of the 1 H NMR spectra between 5a and 6 clearly indicated the disappearance of methine 3a-H (δ 3.86-3.91 ppm) and methylene 3-H protons (δ 4.66 and 3.79 ppm) and the formation of a new downfield methine 3-H proton signal at δ 8.21 ppm. The aforementioned methine proton that appeared as a triplet was mutually coupled with methylene 4-H protons (doublet, δ 5.41 ppm), as indicated by their meta-coupling ( 4 J HH = 1.3 Hz). Moreover, a comparison between the 1 H-1 H COSY and 1 H-1 H NOESY spectra showed a complete absence of COSY cross-peaks between 3-H and 4-H and only strong NOEs, which confirmed their proximity in space. This finding strongly hinted at a neighboring quaternary carbon at site 3a, which was unambiguously assigned from 1,1-ADEQUATE spectral data, where the protonated carbons C-3 (δ 150.7 ppm) and C-4 (δ 63.3 ppm) showed a sole correlation with C-3a at δ 109.8 ppm. As expected, the 15 N chemical shifts of N-6 (δ −116.3) and N-7 (δ −179.6) atoms were highly comparable to those of compounds 5a-c; only the N-1 atoms were slightly different and resonated at δ −20.4 ppm, which is in good agreement with the data reported in the literature [60].
Molecules 2021, 26, 5604 6 of 18 benzyloxypyrazole was formylated under the Vilsmeier-Haack reaction conditions, and the protecting OBn group was cleaved by TFA to give 3-hydroxy-1H-pyrazole-4-carbaldehyde 7 [35]. The latter compound was subjected to an alkylation reaction with cinnamyl chloride and the appropriate 3-cinnamyloxy-1H-pyrazole-4-carbaldehyde (8)  While the structural elucidation of compound trans-10 was straightforward and followed the same logical approach as in the case of compounds 5a-d and 6, determination of the relative configuration at C-3 and C-3a proved to be a more challenging task and was achieved by combined analysis of NOESY, J-coupling and molecular modeling data. For instance, the initial geometry optimizations were performed using MM2 and MMFF94 force fields [61], followed by DFT methods using B3LYP/def2-TZVP, as implemented in ORCA 5.0.0 [62], which provided the dihedral angle values between H-C(3)-C(3a)-H for structures trans-10 (154.34°) and cis-10 (19.28°). Then, the theoretical 1 H-1 H coupling constants were calculated with the same software package following a standard procedure using a B3LYP/PCSSEG-2 basis set. The dihedral angle values were used in the calculation of 3 JH3,H3a by the Haasnoot-de Leeuw-Altona (HLA) equation [63]. The 3 JH3,H3a values estimated by the HLA method were 10.0 Hz for trans-10 and 8.2 Hz for cis-10, while ORCA 5.0.0 calculations were 13.4 and 10.8 Hz, respectively. The experimental value 13.1 Hz, which was obtained from the 1 H NMR spectrum, hinted in favor of the trans-10 structure. A highly similar class of heterocycles, naphthopyranoisoxazolines, were synthesized by Liaskopoulos et al. [64], where the target compounds possessed a trans configuration, as confirmed by X-ray and NMR analyses, and their 3 JH3,H3a values were in the range of 12.2-12.5 Hz. Finally, unambiguous confirmation of trans-10 assignment was obtained from the 1 H-1 H NOESY spectrum, as it was evident from the geometrically optimized structures ( Figures S80 and S81) that, in the case of cis-10, there should be a strong NOE between protons 3-H and 3a-H, while the NOE between 3a-H and the neighboring 3-phenyl group aromatic protons is not possible. However, in our case, the 1 H-1 H NOESY spectrum While the structural elucidation of compound trans-10 was straightforward and followed the same logical approach as in the case of compounds 5a-d and 6, determination of the relative configuration at C-3 and C-3a proved to be a more challenging task and was achieved by combined analysis of NOESY, J-coupling and molecular modeling data. For instance, the initial geometry optimizations were performed using MM2 and MMFF94 force fields [61], followed by DFT methods using B3LYP/def2-TZVP, as implemented in ORCA 5.0.0 [62], which provided the dihedral angle values between H-C(3)-C(3a)-H for structures trans-10 (154.34 • ) and cis-10 (19.28 • ). Then, the theoretical 1 H-1 H coupling constants were calculated with the same software package following a standard procedure using a B3LYP/PCSSEG-2 basis set. The dihedral angle values were used in the calculation of 3 J H3,H3a by the Haasnoot-de Leeuw-Altona (HLA) equation [63]. The 3 J H3,H3a values estimated by the HLA method were 10.0 Hz for trans-10 and 8.2 Hz for cis-10, while ORCA 5.0.0 calculations were 13.4 and 10.8 Hz, respectively. The experimental value 13.1 Hz, which was obtained from the 1 H NMR spectrum, hinted in favor of the trans-10 structure. A highly similar class of heterocycles, naphthopyranoisoxazolines, were synthesized by Liaskopoulos et al. [64], where the target compounds possessed a trans configuration, as confirmed by X-ray and NMR analyses, and their 3 J H3,H3a values were in the range of 12.2-12.5 Hz. Finally, unambiguous confirmation of trans-10 assignment was obtained from the 1 H-1 H NOESY spectrum, as it was evident from the geometrically optimized structures (Figures S80 and S81) that, in the case of cis-10, there should be a strong NOE between protons 3-H and 3a-H, while the NOE between 3a-H and the neighboring 3-phenyl group aromatic protons is not possible. However, in our case, the 1 H-1 H NOESY spectrum showed completely opposite measurements. Moreover, a distinct NOE between protons 3-H/4-H a and 3a-H/4-H b is only possible if the relative configuration is trans-10.
We also investigated the INOC reaction of vic-alkyne-oxime substrates 12 and 14a-c (Scheme 3). To obtain the intermediate compound 12, firstly, 3-hydroxypyrazole 1a was O-propargylated and formylated to give carbaldehyde 11 [26]. Compound 11 was then successfully converted to 4H,7H-pyrazolo [4 ,3 :5,6]pyrano [4,3-c] [1,2]oxazole 6 via the INOC reaction of intermediate oxime 12, and the targeted new polyheterocyclic compound 6 was obtained in good (79%) yield. In addition, alkyne 11 was further subjected to the Sonogashira cross-coupling reaction with various (het)arylhalides, i.e., iodobenzene, 1iodonaphthalene and 2-bromopyridine, under the standard Sonogashira cross-coupling reaction conditions (Pd(PPh 3 ) 2 Cl 2 , CuI, DMF, 60 • C, argon atmosphere) to give alkynes 13a-c in good yields [26]. Compounds 13a-c were further treated with hydroxylamine hydrochloride to provide aldoximes 14a-c, which were used in the INOC reaction without further purification. Aldoxime 14a was subjected to a detailed NMR analysis, and, to our delight, it was obtained as a pure syn isomer, which was easily elucidated from a 1 J CH coupling constant of the iminyl moiety, which was 179. , and the targeted new polyheterocyclic compound 6 was obtained in good (79%) yield. In addition, alkyne 11 was further subjected to the Sonogashira cross-coupling reaction with various (het)arylhalides, i.e., iodobenzene, 1-iodonaphthalene and 2-bromopyridine, under the standard Sonogashira crosscoupling reaction conditions (Pd(PPh3)2Cl2, CuI, DMF, 60 °C, argon atmosphere) to give alkynes 13a-c in good yields [26]. Compounds 13a-c were further treated with hydroxylamine hydrochloride to provide aldoximes 14a-c, which were used in the INOC reaction without further purification. Aldoxime 14a was subjected to a detailed NMR analysis, and, to our delight, it was obtained as a pure syn isomer, which was easily elucidated from a 1 JCH coupling constant of the iminyl moiety, which was 179.2 Hz. Moreover, 4H,7H-pyrazolo[4′,3′:5,6]pyrano [4,3-c] [1,2]oxazoles 15a-c were obtained in good yields. As expected, the chemical shifts of the 3-aryl-substituted compounds 15a-c were highly similar to those of compound 6. A distinct difference in the 1 H NMR spectra of the aforementioned compounds was that they contained only a singlet for the methylene 4-H protons in the area of δ 5.32-6.03 ppm, which indicated the lack of coupling partners. The data from the 1 H-13 C HMBC spectra revealed a distinct long-range correlation between the aforementioned methylene protons and a quaternary carbon at site 3. Moreover, the protons from a neighboring 3-aryl moiety shared an HMBC cross-peak with carbon C-3 as well, thus allowing different structural fragments to be joined together. The chemical shifts of the N-1, N-6 and N-7 atoms of 3-aryl-substituted compounds were in ranges of δ −23.9 to −25.0, δ −116.4 to −117.4 and δ −179.6 to −180.1 ppm, respectively, while, in the case of compound 15c with a pyridin-2-yl moiety, the pyridine nitrogen resonated at δ −72.8 ppm. As expected, the chemical shifts of the 3-aryl-substituted compounds 15a-c were highly similar to those of compound 6. A distinct difference in the 1 H NMR spectra of the aforementioned compounds was that they contained only a singlet for the methylene 4-H protons in the area of δ 5.32-6.03 ppm, which indicated the lack of coupling partners. The data from the 1 H-13 C HMBC spectra revealed a distinct long-range correlation between the aforementioned methylene protons and a quaternary carbon at site 3. Moreover, the protons from a neighboring 3-aryl moiety shared an HMBC cross-peak with carbon C-3 as well, thus allowing different structural fragments to be joined together. The chemical shifts of the N-1, N-6 and N-7 atoms of 3-aryl-substituted compounds were in ranges of δ −23.9 to −25.0, δ −116.4 to −117.4 and δ −179.6 to −180.1 ppm, respectively, while, in the case of compound 15c with a pyridin-2-yl moiety, the pyridine nitrogen resonated at δ −72.8 ppm.

General Information
All starting materials were purchased from commercial suppliers and were used as received. Flash column chromatography was performed on Silica Gel 60 Å (230-400 µm, Merck). Thin-layer chromatography was carried out on Silica Gel plates (Merck Kieselgel 60 F 254 ) and visualized by UV light (254 nm). Melting points were determined on a Büchi M-565 melting point apparatus and were uncorrected. The IR spectra were recorded on a Bruker Vertex 70v FT-IR spectrometer using neat samples and are reported in frequency of absorption (cm −1 ). Mass spectra were obtained on a Shimadzu LCMS-2020 (ESI + ) spectrometer. High-resolution mass spectra were measured on a Bruker MicrOTOF-Q III (ESI + ) apparatus. The 1 H, 13 C and 15 N NMR spectra were recorded in CDCl 3 , DMSO-d 6 or TFA-d solutions at 25 • C on a Bruker Avance III 700 (700 MHz for 1 H, 176 MHz for 13 C, 71 MHz for 15 N) spectrometer equipped with a 5 mm TCI 1 H-13 C/ 15 N/D z-gradient cryoprobe. The chemical shifts (δ), expressed in ppm, were relative to tetramethylsilane (TMS). The 15 N NMR spectra were referenced to neat, external nitromethane (coaxial capillary). Full and unambiguous assignment of the 1 H, 13 C and 15 N NMR resonances was achieved using a combination of standard NMR spectroscopic techniques [65] such as DEPT, COSY, TOCSY, NOESY, gs-HSQC, gs-HMBC, H2BC, LR-HSQMBC and 1,1-ADEQUATE experiments [66]. Structures for molecular modeling were built using Chem3D Pro 17.0, and were optimized by MM2 and MMFF94 force fields, followed by DFT methods using B3LYP/def2-TZVP for dihedral angle measurements, and B3LYP/PCSSEG-2 for the calculation of theoretical 1 H-1 H coupling constants, using a standard procedure as implemented in the ORCA 5.

Synthesis of 1-Methyl-3-[(prop-2-en-1-yl)oxy]-1H-pyrazole-4-carbaldehyde (3d)
Phosphorus oxychloride (0.2 mL, 2.5 mmol) was added dropwise to DMF (0.23 mL, 2.5 mmol) at −10 • C temperature. Then, pyrazole 2d (0.62 mmol) was added to the Vilsmeier-Haack complex, and the reaction mixture was heated at 70 • C temperature for 1 h. After neutralization with 10% NaHCO 3 solution, it was extracted with ethyl acetate (3 × 10 mL). The organic layers were combined, washed with brine, dried over Na 2 SO 4 , filtrated, and the solvent was evaporated. The residue was purified by flash column chromatography (SiO 2 , eluent: ethyl acetate/n-hexane, 1:6, v/v) to provide the desired compound 3c as a brown solid, yield 400 mg, 95%, mp 59-60 • C. IR (KBr, v max , cm −1 ): 3122,  (3 mmol) in EtOH (10 mL), sodium acetate (369 mg, 4.5 mmol) and hydroxylamine hydrochloride (250 mg, 3.6 mmol) were added, and the reaction mixture was refluxed for 15 min. After completion of the reaction as monitored by TLC, EtOH was evaporated, and the mixture was diluted with water (10 mL) and extracted with ethyl acetate (3 × 10 mL). The organic layers were combined, washed with brine, dried over Na 2 SO 4 , filtrated, and the solvent was evaporated. The residue was purified by flash column chromatography (SiO 2 , eluent: ethyl acetate/n-hexane, 1:6, v/v) to provide the desired compounds 4a-c, as mixtures of syn-anti isomers or pure syn isomer 4d. Due to small extent of minor isomer and heavy overlap with signals of the predominant syn-(Z) isomer, NMR spectroscopy data of the major isomer only are presented, while the relevant NMR spectroscopy data of the minor isomer are presented in a supplementary file, Table S1.

General Procedure for the Cycloaddition Reaction of Pyrazole Oximes 4a-d
Into the solution of appropriate pyrazole (0.4 mmol) 4a-d in DCM (5 mL), sodium hypochlorite (10% aq. solution, 0.5 mL, 0.8 mmol) was added, and the reaction mixture was stirred for 1 h at room temperature. After completion of the reaction as monitored by TLC, it was diluted with water (10 mL) and extracted with DCM (3 × 10 mL). The organic layers were combined, washed with brine, dried over Na 2 SO 4 , filtrated, and the solvent was evaporated. The residue was purified by flash column chromatography (SiO 2 , eluent: ethyl acetate/n-hexane, 1:4, v/v) to provide the desired compounds 5a-d.