Palladium-Catalyzed Domino Cycloisomerization/Double Condensation of Acetylenic Acids with Dinucleophiles

: Metal-catalyzed cascade processes based on hydrofunctionalization of alkynes are receiving much more attention because of their potential to provide advantageous approaches to otherwise synthetically challenging compounds. An alternative catalyst system has been found for the domino cycloisomerization/cyclocondensation reaction involving acetylenic acids and heterodinucleophiles. A CNN pincer palladium(II) complex, acting as a homogeneous catalyst, provides the corresponding polyheterocycles with a higher substrate/catalyst ratio. Other palladium sources were also tested and discarded, and a number of mechanistic studies including poisoning assays, kinetic plots, TEM images, XRD spectra and UPLC-MS analysis of reaction intermediates were conducted in order to shed light on the role of this pincer catalyst and the catalytic cycle involved in the cascade reaction. As a result, a more nuanced mechanism is tentatively proposed.

Following our research on metal-catalyzed transformations involving alkynes [23], we had described an unprecedented amount of activity for the cycloisomerization of alkynoic acids exhibited by a palladium NNC complex (1) [24]. On this basis, we envisaged that such a more active catalyst could also promote a domino process leading to polyheterocyclic structures if reacted with suitable dinucleophiles. In addition to a new metal participating in this reaction, this approach could lead to a substantial enhancement in the catalytic efficiency, which is increasingly becoming a priority in sustainable chemistry [25][26][27]. In this regard, we wish to present an exceedingly active catalyst system for the cascade Following our research on metal-catalyzed transformations involving alkynes [23], we had described an unprecedented amount of activity for the cycloisomerization of alkynoic acids exhibited by a palladium NNC complex (1) [24]. On this basis, we envisaged that such a more active catalyst could also promote a domino process leading to polyheterocyclic structures if reacted with suitable dinucleophiles. In addition to a new metal participating in this reaction, this approach could lead to a substantial enhancement in the catalytic efficiency, which is increasingly becoming a priority in sustainable chemistry [25][26][27]. In this regard, we wish to present an exceedingly active catalyst system for the cascade reaction between alkynoic acids and N,O-and N,N-dinucleophiles as well as an insight into the reaction mechanism.

Results and Discussion
Pentynoic acid 2a and anthranilic acid 3a were chosen as model substrates for a number of initial assays. Different palladium sources (Pd(OAc)2, PdCl2, and Pd(PPh3)4) were tested, and the results compared to those obtained from NNC palladium complex 1. The catalyst loading was set at 10 −2 mol% in order to meet the aforementioned requirements  Following our research on metal-catalyzed transformations involving alkynes [23], we had described an unprecedented amount of activity for the cycloisomerization of alkynoic acids exhibited by a palladium NNC complex (1) [24]. On this basis, we envisaged that such a more active catalyst could also promote a domino process leading to polyheterocyclic structures if reacted with suitable dinucleophiles. In addition to a new metal participating in this reaction, this approach could lead to a substantial enhancement in the catalytic efficiency, which is increasingly becoming a priority in sustainable chemistry [25][26][27]. In this regard, we wish to present an exceedingly active catalyst system for the cascade reaction between alkynoic acids and N,O-and N,N-dinucleophiles as well as an insight into the reaction mechanism.

Results and Discussion
Pentynoic acid 2a and anthranilic acid 3a were chosen as model substrates for a number of initial assays. Different palladium sources (Pd(OAc)2, PdCl2, and Pd(PPh3)4) were tested, and the results compared to those obtained from NNC palladium complex 1. The catalyst loading was set at 10 −2 mol% in order to meet the aforementioned requirements Scheme 1. Postulated catalytic cycle and reported metal catalysts for the cascade reaction between alkynoic acids and dinucleophiles.

Results and Discussion
Pentynoic acid 2a and anthranilic acid 3a were chosen as model substrates for a number of initial assays. Different palladium sources (Pd(OAc) 2 , PdCl 2 , and Pd(PPh 3 ) 4 ) were tested, and the results compared to those obtained from NNC palladium complex 1. The catalyst loading was set at 10 −2 mol% in order to meet the aforementioned requirements of efficiency. The amounts of base and/or additive were also minimized following the same criteria. After a number of initial experiments, it became clear that the process required chloroform as a solvent and triethylamine as a base to get reasonable yields, as negligible results were obtained from other bases (KOH, NaOH, LiOH, KO t Bu, K 2 CO 3 , Cs 2 CO 3 , DBU, DIPEA, pyridine and DMAP) and solvents (THF, PhMe, MeOH, CH 2 Cl 2 , DMF, DMA, DMSO and MeCN) (see Supplementary Material for further details).
Only pincer complex 1 provided target benzopyrrolo [1,3]oxazine-1,5-dione 4a even using triethylamine in chloroform ( Figure 2, entries 1-4), and the presence of a Lewis acid additive proved beneficial for the reaction outcome (entries 5-9). With a catalytic amount (0.01 mol%) of iron(II) bromide as the most favorable additive, an increase in the reaction temperature (120 °C) led to the desired product with a good yield (79%, entry 10). We hypothesized that both the iron Lewis acid and higher temperature favored the key dehydration step. Attempts to decrease the amounts of both the catalyst and the additive (entries 11,12), and to improve the yield by adding a little With a catalytic amount (0.01 mol%) of iron(II) bromide as the most favorable additive, an increase in the reaction temperature (120 • C) led to the desired product with a good yield (79%, entry 10). We hypothesized that both the iron Lewis acid and higher temperature favored the key dehydration step. Attempts to decrease the amounts of both the catalyst and the additive (entries 11,12), and to improve the yield by adding a little excess of the alkynoic acid or changing the reaction concentration (entries [13][14][15][16][17][18] were made so that an excellent 93% yield was obtained by the optimized conditions of entry 17. Lower temperatures or shorter reaction times (entries 19,20) provided poorer results, and a series of blank experiments proved the need of all the ingredients, including Et 3 N as a base and FeBr 2 as an additive, along with the palladium source (entries 21-23 vs. 17). In this regard, we were surprised to find that target 4a was isolated with a yield of 86% when complex 1 was replaced with Pd(OAc) 2 (entry 28). Therefore, when the optimized conditions ( Figure 2, entry 21) were applied to a number of commercially or readily available (Supplementary Material) alkynoic acids 2 and heterodinucleophiles 3, Pd(OAc) 2 was also evaluated as a catalyst for the reaction.
when complex 1 was replaced with Pd(OAc)2 (entry 28). Therefore, when the optimized conditions ( Figure 2, entry 21) were applied to a number of commercially or readily available (Supplementary Material) alkynoic acids 2 and heterodinucleophiles 3, Pd(OAc)2 was also evaluated as a catalyst for the reaction.
Structure analysis of monocrystals of 4a and 4e revealed that both have the same crystal system and space group (orthorrombic Pbca). The presence of the C3a quaternary center (C-11) is responsible for the lack of coplanarity of oxazine and pyrrole rings, as shown by the C1-O1-C11-N1 and C7-N1-C11-C10 torsion angle values of −53.4-39.6 • and 165.1-164.4 • , respectively. Only the R enantiomer of sultam 4g crystallized in the monoclinic system (Cc) leaving the other C-5 isomer in an amorphous state ( Figure 4). the first benzopyrido [1,2,4]thiadiazinone-5,5-dioxide synthesized so far.
Structure analysis of monocrystals of 4a and 4e revealed that both have the same crystal system and space group (orthorrombic Pbca). The presence of the C3a quaternary center (C-11) is responsible for the lack of coplanarity of oxazine and pyrrole rings, as shown by the C1-O1-C11-N1 and C7-N1-C11-C10 torsion angle values of −53.4-39.6° and 165.1-164.4°, respectively. Only the R enantiomer of sultam 4g crystallized in the monoclinic system (Cc) leaving the other C-5 isomer in an amorphous state ( Figure 4). Several diamines 5 were also reacted with alkynoic acids 2. Palladacycle 1 catalyzed cascade process on a general basis, whereas Pd(OAc)2 did not provide most of the target tri-and tetracycles 6. In fact, only quinazolinones 6a and 6c were obtained when using the latter commercially available palladium source, although with much inferior yields (Figure 5).
As in the case of derivatives 4, the reason probably lies in the inability of palladium(II) acetate to catalyze the initial cycloisomerization step at such low catalyst levels. By using remarkably active palladacycle 1, pyrrolo-, isoindolo-and pyrido[2,1-b]quinazolinones 6a-d were regio-selectively prepared from o-aminobenzylamine 5a. As in the case of anthranilic acid 3a or 2-aminobenzenesulfonamide 3b, the use of internal alkynes did not hinder the reaction, and 3a-benzyl derivative 6d was obtained in good yield. Besides, reaction of 5a with 5-hexynoic acid provided pyrido[2,1-b]quinazolin-9-one 6c. So far, this compound had been prepared in reasonable yield only from 5-oxohexanoic acid by a Meyers' lactamization [38].
Several diamines 5 were also reacted with alkynoic acids 2. Palladacycle 1 catalyzed cascade process on a general basis, whereas Pd(OAc) 2 did not provide most of the target tri-and tetracycles 6. In fact, only quinazolinones 6a and 6c were obtained when using the latter commercially available palladium source, although with much inferior yields ( Figure 5). Several experiments were then carried out in order to shed light on the role of complex 1 in the reaction mechanism. Taking the reaction leading to the formation of benzopyrrolo [1,3]oxazine-1,5-dione 4a as a model example, a plot of the conversion of anthranilic acid 3a vs. time showed neither sigmoidal shape nor induction time ( Figure 6). As in the case of derivatives 4, the reason probably lies in the inability of palladium(II) acetate to catalyze the initial cycloisomerization step at such low catalyst levels. By using remarkably active palladacycle 1, pyrrolo-, isoindolo-and pyrido[2,1-b]quinazolinones 6a-d were regio-selectively prepared from o-aminobenzylamine 5a. As in the case of anthranilic acid 3a or 2-aminobenzenesulfonamide 3b, the use of internal alkynes did not hinder the reaction, and 3a-benzyl derivative 6d was obtained in good yield. Besides, Catalysts 2022, 12, 127 6 of 14 reaction of 5a with 5-hexynoic acid provided pyrido[2,1-b]quinazolin-9-one 6c. So far, this compound had been prepared in reasonable yield only from 5-oxohexanoic acid by a Meyers' lactamization [38].
Several experiments were then carried out in order to shed light on the role of complex 1 in the reaction mechanism. Taking the reaction leading to the formation of benzopyrrolo [1,3]oxazine-1,5-dione 4a as a model example, a plot of the conversion of anthranilic acid 3a vs. time showed neither sigmoidal shape nor induction time ( Figure 6). Several experiments were then carried out in order to shed light on the role of complex 1 in the reaction mechanism. Taking the reaction leading to the formation of benzopyrrolo [1,3]oxazine-1,5-dione 4a as a model example, a plot of the conversion of anthranilic acid 3a vs. time showed neither sigmoidal shape nor induction time ( Figure 6). In addition, the latter reaction was also conducted in the presence of substoichiometric and overstoichiometric amounts of a number of known poisoning agents for palladium nanoparticles and other palladium(0) heterogeneous catalysts, including the mercury drop test (Table 1). Not the slightest inhibition or poisoning effect was observed in our case, as expected for a fully homogeneous catalyst system [42][43][44][45][46][47]. In addition, the latter reaction was also conducted in the presence of substoichiometric and overstoichiometric amounts of a number of known poisoning agents for palladium nanoparticles and other palladium(0) heterogeneous catalysts, including the mercury drop test (Table 1). Not the slightest inhibition or poisoning effect was observed in our case, as expected for a fully homogeneous catalyst system [42][43][44][45][46][47]. Moreover, transmission electron microscopy (TEM) analysis of the reaction mixture from 3a + 2a -> 4a, combined with energy dispersive X-ray microanalysis revealed the presence of some scattered iron nanoparticles (average size 8 nm) but no trace of palladium nanoparticles (Figure 7). Moreover, transmission electron microscopy (TEM) analysis of the reaction mixture from 3a +2a ≥ 4a, combined with energy dispersive X-ray microanalysis revealed the presence of some scattered iron nanoparticles (average size 8 nm) but no trace of palladium nanoparticles (Figure 7).  UPLC-ESI analysis of a sample taken from the reaction mixture (2a + 3a -> 4a) after 12 h of reaction not only provided evidence of the presence of several transient intermediates related to the ones postulated in Scheme 1, but also of the participation of a number of palladium and iron species displayed in Scheme 2 (see Supplementary Material for further details), thus supporting our hypothesis on the crucial role of such ingredients in the cascade reaction. On the basis of the detection of the above intermediates, a revised catalytic cycle was tentatively proposed (Scheme 2). Interaction of alkynoic acid 2a with cationic 1 to form complex A activates the alkyne for the triethylamine-promoted cycloisomerization process that generates intermediate B, which upon protonation and interaction with Lewis's acids 1 and FeBr 2 produces transient species C. Ring-opening amidation with o-amino derivative 3a leads to activated ketoamides D, which upon cyclization form hydroxypyrrolidinones E. After dehydration, the resulting N-acyliminium intermediates F undergo a final lactamization step to provide oxazinanone derivative 4a. According to the transient species detected by ESI-mass spectrometry (Supplementary Material), it seems that although the palladium catalyst is essential to initiate the cycloisomerization step and thus to promote the reaction, the iron additive facilitates several nucleophilic addition and dehydration steps via Lewis acid activation.
Finally, methylidene lactone 7 was prepared using catalyst 1, and then reacted with anthranilic acid 3a in the presence and absence of the optimized catalyst and additives. Target benzopyrrolo [1,3]oxazine-1,5-dione 4a was obtained in good yield (90%, Scheme 3) only when both complex 1 and FeBr 2 were employed, whereas reactions lacking any of these ingredients led to negligible results (yield of 4a < 15%) (see Supplementary Material for further details), thus offering further support for the proposed reaction pathway. with o-amino derivative 3a leads to activated ketoamides D, which upon cyclization form hydroxypyrrolidinones E. After dehydration, the resulting N-acyliminium intermediates F undergo a final lactamization step to provide oxazinanone derivative 4a. According to the transient species detected by ESI-mass spectrometry (Supplementary Material), it seems that although the palladium catalyst is essential to initiate the cycloisomerization step and thus to promote the reaction, the iron additive facilitates several nucleophilic addition and dehydration steps via Lewis acid activation.

Scheme 2.
Revised catalytic cycle for the cascade reaction catalyzed by the 1/FeBr3/Et3N system. Finally, methylidene lactone 7 was prepared using catalyst 1, and then reacted with anthranilic acid 3a in the presence and absence of the optimized catalyst and additives. Target benzopyrrolo [1,3]oxazine-1,5-dione 4a was obtained in good yield (90%, Scheme 3) only when both complex 1 and FeBr2 were employed, whereas reactions lacking any of these ingredients led to negligible results (yield of 4a <15%) (see Supplementary Material for further details), thus offering further support for the proposed reaction pathway.

General Information
Commercially available reagents were used throughout without purification unless otherwise stated. 1 H and 13 C NMR spectra were recorded on a Bruker AC-300 instrument (300 MHz for 1 H and 75.4 MHz for 13 C) at 20 °C. Chemical shifts (d) are given in ppm downfield from Me4Si and are referenced as internal standard to the residual solvent Scheme 2. Revised catalytic cycle for the cascade reaction catalyzed by the 1/FeBr 3 /Et 3 N system.

Scheme 2.
Revised catalytic cycle for the cascade reaction catalyzed by the 1/FeBr3/Et3N system. Finally, methylidene lactone 7 was prepared using catalyst 1, and then reacted with anthranilic acid 3a in the presence and absence of the optimized catalyst and additives. Target benzopyrrolo [1,3]oxazine-1,5-dione 4a was obtained in good yield (90%, Scheme 3) only when both complex 1 and FeBr2 were employed, whereas reactions lacking any of these ingredients led to negligible results (yield of 4a <15%) (see Supplementary Material for further details), thus offering further support for the proposed reaction pathway.

General Information
Commercially available reagents were used throughout without purification unless otherwise stated. 1 H and 13 C NMR spectra were recorded on a Bruker AC-300 instrument (300 MHz for 1 H and 75.4 MHz for 13 C) at 20 °C. Chemical shifts (d) are given in ppm downfield from Me4Si and are referenced as internal standard to the residual solvent

General Information
Commercially available reagents were used throughout without purification unless otherwise stated. 1 H and 13 C NMR spectra were recorded on a Bruker AC-300 instrument (300 MHz for 1 H and 75.4 MHz for 13 C) at 20 • C. Chemical shifts (d) are given in ppm downfield from Me 4 Si and are referenced as internal standard to the residual solvent (unless indicated) CDCl 3 (d = 7.26 for 1 H and d = 77.00 for 13 C). Coupling constants, J, are reported in hertz (Hz). Melting points were determined in a capillary tube and are uncorrected. TLC was carried out on SiO 2 (silica gel 60 F254, Merck), and the spots were located with UV light. Flash chromatography was carried out on SiO 2 (silica gel 60, Merck, 230-400 mesh ASTM). Drying of organic extracts during work-up of reactions was performed over anhydrous Na 2 SO 4 . Evaporation of solvents was accomplished with a Büchi rotatory evaporator. MS spectra were recorded on an Agilent 5975 mass spectrometer under electronic impact (EI) conditions or on an Acquity UPLC-Spectrometer of Mass QTOF from Waters under electrospray ionization (ESI). HRMS were recorded using a Micromass GCT spectrometer by electronic impact (EI) or electrospray ionization (ESI). Single crystal X-ray intensity data were collected on an Agilent Technologies Super-Nova diffractometer, which was equipped with monochromated Cu kα radiation (λ = 1.54184 A) and Atlas CCD detector. A measurement was carried out at 100(2) K with the help of an Oxford Cryostream 700 PLUS temperature device. Data frames were processed (united cell determination, analytical absorption correction with face indexing, intensity data integration and correction for Lorentz and polarization effects) using the Crysalis software package [48]. The structure was solved using Olex2 [49] and refined by full-matrix least-squares with SHELXL-97 [50].