Eosin Y-Catalyzed Visible-Light-Mediated Aerobic Transformation of Pyrazolidine-3-One Derivatives †

: By utilizing an underexplored reactivity of N1-substituted pyrazolidine-3-ones, we developed a visible-light-induced aerobic oxidation of N1-substituted pyrazolidine-3-one derivatives yielding the corresponding azomethine imines. The resulting azomethine imines can be further reacted with ynones in situ under copper catalyzed [3 + 2] cycloaddition reaction conditions yielding the corresponding pyrazolo[1,2- a ]pyrazoles in good yields. The methodology can be extended to other pyrazolidinones which undergo endocyclic oxidation deriving the corresponding pyrazolones as single products.

The use of visible-light-mediated photoredox catalysis as an environmentally friendly, sustainable, effective, and versatile method for the functionalization of a variety of systems such as relatively unreactive C-H bonds adjacent to nitrogen or oxygen atoms, has been developed. The pioneering work in this research area, reported by the research groups of MacMillan [23][24][25], Yoon [26,27], Stephenson [28,29], and others [30][31][32][33][34], has shown that particular Ru and Ir complexes can act as visible-light photoredox catalysts capable of catalyzing a broad range of organic reactions. In contrast, other metal complexes such as Cu-based photocatalysts are emerging, offering not only economic and ecological advantages but also unique reactivity modes [35,36]. The use of organic dyes as photoredox catalysts [37] has been demonstrated to be a useful alternative to the inorganic photoredox catalysts that are expensive and sometimes toxic. Small organic molecules have also shown their activity towards the activation of the O 2 molecule under visible light, and this synthetic strategy has already shown powerful applications in organic synthesis via C-H bond activation and others [38][39][40][41]. Moreover, from an environmental point of view, oxygen (O 2 and air) represents an oxidant of choice, and its advantages include abundance, low cost, and production of water as a by-product [42,43].
Photooxidation of amine functionalities to amine-centered radical cations, displaying several modes of reactivity (Scheme 1A), has become a major approach to exploit the synthetic utility of photogenically produced amine radical cations [44]. As a part of our ongoing research on the reactivity and stereoselectivity of azomethine imines we hypothesized that N1-substituted pyrazolidine-3-ones could derive the azomethine imine functionality upon photooxidative conditions (Scheme 1B).
Catalysts 2020, 10, x FOR PEER REVIEW 2 of 10 [26,27], Stephenson [28,29], and others [30][31][32][33][34], has shown that particular Ru and Ir complexes can act as visible-light photoredox catalysts capable of catalyzing a broad range of organic reactions. In contrast, other metal complexes such as Cu-based photocatalysts are emerging, offering not only economic and ecological advantages but also unique reactivity modes [35,36]. The use of organic dyes as photoredox catalysts [37] has been demonstrated to be a useful alternative to the inorganic photoredox catalysts that are expensive and sometimes toxic. Small organic molecules have also shown their activity towards the activation of the O2 molecule under visible light, and this synthetic strategy has already shown powerful applications in organic synthesis via C-H bond activation and others [38][39][40][41]. Moreover, from an environmental point of view, oxygen (O2 and air) represents an oxidant of choice, and its advantages include abundance, low cost, and production of water as a by-product [42,43]. Photooxidation of amine functionalities to amine-centered radical cations, displaying several modes of reactivity (Scheme 1A), has become a major approach to exploit the synthetic utility of photogenically produced amine radical cations [44]. As a part of our ongoing research on the reactivity and stereoselectivity of azomethine imines we hypothesized that N1-substituted pyrazolidine-3-ones could derive the azomethine imine functionality upon photooxidative conditions (Scheme 1B). Scheme 1. Visible-light-induced sp 3 C-H bond functionalization of α-amino alkyl radical (A) and schematic representation of the transformation reported herein (B).

Results and Discussion
Herein, the development of visible-light-mediated aerobic transformation of N1-substituted pyrazolidine-3-one 1 derivatives yielding the corresponding azomethine imines 2 or pyrazolo[1,2-a]pyrazoles 3 and alternatively N1-substituted 1,2-dihydro-3H-pyrazol-3-ones 5 is disclosed (Scheme 2). Molecular oxygen or air serves as a sacrificial green and abounded oxidant in combination with eosin Y as the photocatalyst under blue LED irradiation. In the initial experiments, transformation of 1-(4-chlorobenzyl)-5,5-dimethylpirazolidin-3-one (1c) was conducted under different reaction conditions to yield the corresponding azomethine imine 2c. Running the reaction in acetonitrile in the presence of 5 mol% of eosin Y, air, and trifluoroacetic acid (TFA, 2 equiv.) under LED irradiation (470 nm) for 60 min resulted in the formation of 2c in a 62% isolated yield (Table 1, entry 1). Optimization of the reaction conditions revealed, that addition of trifluoroacetic acid (TFA) was essential to achieve high conversions of 1c to 2c and to slow down the overoxidation of 2c, yielding the corresponding 4-chlorobenzaldehyde as a side product (Table 1,
Catalysts 2020, 10, x FOR PEER REVIEW 2 of 10 [26,27], Stephenson [28,29], and others [30][31][32][33][34], has shown that particular Ru and Ir complexes can act as visible-light photoredox catalysts capable of catalyzing a broad range of organic reactions. In contrast, other metal complexes such as Cu-based photocatalysts are emerging, offering not only economic and ecological advantages but also unique reactivity modes [35,36]. The use of organic dyes as photoredox catalysts [37] has been demonstrated to be a useful alternative to the inorganic photoredox catalysts that are expensive and sometimes toxic. Small organic molecules have also shown their activity towards the activation of the O2 molecule under visible light, and this synthetic strategy has already shown powerful applications in organic synthesis via C-H bond activation and others [38][39][40][41]. Moreover, from an environmental point of view, oxygen (O2 and air) represents an oxidant of choice, and its advantages include abundance, low cost, and production of water as a by-product [42,43]. Photooxidation of amine functionalities to amine-centered radical cations, displaying several modes of reactivity (Scheme 1A), has become a major approach to exploit the synthetic utility of photogenically produced amine radical cations [44]. As a part of our ongoing research on the reactivity and stereoselectivity of azomethine imines we hypothesized that N1-substituted pyrazolidine-3-ones could derive the azomethine imine functionality upon photooxidative conditions (Scheme 1B).

Results and Discussion
Herein, the development of visible-light-mediated aerobic transformation of N1-substituted pyrazolidine-3-one 1 derivatives yielding the corresponding azomethine imines 2 or pyrazolo[1,2-a]pyrazoles 3 and alternatively N1-substituted 1,2-dihydro-3H-pyrazol-3-ones 5 is disclosed (Scheme 2). Molecular oxygen or air serves as a sacrificial green and abounded oxidant in combination with eosin Y as the photocatalyst under blue LED irradiation. In the initial experiments, transformation of 1-(4-chlorobenzyl)-5,5-dimethylpirazolidin-3-one (1c) was conducted under different reaction conditions to yield the corresponding azomethine imine 2c. Running the reaction in acetonitrile in the presence of 5 mol% of eosin Y, air, and trifluoroacetic acid (TFA, 2 equiv.) under LED irradiation (470 nm) for 60 min resulted in the formation of 2c in a 62% isolated yield (Table 1, entry 1). Optimization of the reaction conditions revealed, that addition of trifluoroacetic acid (TFA) was essential to achieve high conversions of 1c to 2c and to slow down the overoxidation of 2c, yielding the corresponding 4-chlorobenzaldehyde as a side product (Table 1,  In the initial experiments, transformation of 1-(4-chlorobenzyl)-5,5-dimethylpirazolidin-3-one (1c) was conducted under different reaction conditions to yield the corresponding azomethine imine 2c. Running the reaction in acetonitrile in the presence of 5 mol% of eosin Y, air, and trifluoroacetic acid (TFA, 2 equiv.) under LED irradiation (470 nm) for 60 min resulted in the formation of 2c in a 62% isolated yield (Table 1, entry 1). Optimization of the reaction conditions revealed, that addition of trifluoroacetic acid (TFA) was essential to achieve high conversions of 1c to 2c and to slow down the overoxidation of 2c, yielding the corresponding 4-chlorobenzaldehyde as a side product (Table 1, entry 13). Additionally, other acids were also employed such as oxalic acid, citric acid, and trichloroacetic acid resulting in lower conversions of the starting material 1c. Trifluoroacetic acid is most probably responsible for the protonation of the N1 nitrogen of azomethine imines 2, which can be easily observed in the 1 H NMR spectra of compounds 2 in CDCl 3 . We presume that the protonated form of 2 is less prone for the photoinduced oxidation yielding the corresponding benzaldehydes under the applied reaction conditions. However, using other solvents than acetonitrile resulted in much lower yields of the corresponding azomethine imine 2c, as determined by 1 H NMR analysis of the crude reaction mixture. When toluene was used as a solvent (Table 1, entry 8) comparable results were observed after 0.5 h of irradiation at room temperature. The prolonged reaction times (2-6 h) result in diminished yields of 2c as 4-chlorobenzaldehyde is being formed as the major product. With the encouragement, the reaction was studied by varying the photocatalyst (Table 1, entries 9-12) however, eosin Y showed superior results and was selected as the most suitable photocatalyst for the transformation being studied. Reducing its loading to 2 mol%, (Table 1, entry 9) marginally decreased the yield of 2c and 5-mol% loading was necessary for short reaction times needed. The somewhat higher catalyst loading may be necessary due to the acidic reaction medium, where eosin Y displays different absorption characteristics. According to the literature [27], Eosin Y is expected to be quite unsuitable as a photocatalyst in acidic medium, however, we found that absorption (at 470 nm) and fluorescence are still present and the absorption maximum moves towards the blue wavelengths (see Supplementary Materials). Increasing the quantity of TFA or dilution of reaction mixture only negligibly improved the yield of the product 2c. However, formation of 2c was observed only in traces when no irradiation was employed or no photocatalyst or oxygen was provided (Table 1, entry 19-21). Table 1. Deviation from the standard reaction conditions for the transformation of pyrazolidine-3-one 1c to azomethine imine 2c.
Catalysts 2020, 10, x FOR PEER REVIEW 3 of 10 13). Additionally, other acids were also employed such as oxalic acid, citric acid, and trichloroacetic acid resulting in lower conversions of the starting material 1c. Trifluoroacetic acid is most probably responsible for the protonation of the N1 nitrogen of azomethine imines 2, which can be easily observed in the 1 H NMR spectra of compounds 2 in CDCl3. We presume that the protonated form of 2 is less prone for the photoinduced oxidation yielding the corresponding benzaldehydes under the applied reaction conditions. However, using other solvents than acetonitrile resulted in much lower yields of the corresponding azomethine imine 2c, as determined by 1 H NMR analysis of the crude reaction mixture. When toluene was used as a solvent ( With the optimal reaction conditions established, starting compound 1c (0.1 mmol), eosin Y-disodium salt (5 mol%), MeCN (1.0 mL), TFA (2.0 equiv.), air, 25-30 • C, 465-470 nm LED, we then examined the substrate scope of the photoinduced oxidation of pyrazolidine-3-ones. As highlighted in Scheme 3, a range of 1-(benzyl)pirazolidin-3-ones were subjected to the optimized reaction conditions yielding the corresponding azomethine imines in good to moderate yields upon isolation. The reaction conditions tolerated a variety of substituents (electron-withdrawing and electron-donating) being present on the aromatic ring (Scheme 3, examples 2a-2h). The C4-aminosubstituted pyrazolidin-3-one furnished the corresponding azomethine imine 2j in a reasonable 48% isolated yield. Structure variation in the benzylidene moiety can also be realized without major loss of reaction efficiency. The incorporation of the heteroaryl functionality such as pyridyl and furyl resulted in the formation of products 2k and 2l in reasonable yields. However, the alkylidene side chain proved to be more challenging. As demonstrated in Scheme 3 azomethine imine 2m was formed in situ with a good 70% yield and was found to be unstable during the isolation. With the optimal reaction conditions established, starting compound 1c (0.1 mmol), eosin Y-disodium salt (5 mol%), MeCN (1.0 mL), TFA (2.0 equiv.), air, 25-30 °C, 465-470 nm LED, we then examined the substrate scope of the photoinduced oxidation of pyrazolidine-3-ones. As highlighted in Scheme 3, a range of 1-(benzyl)pirazolidin-3-ones were subjected to the optimized reaction conditions yielding the corresponding azomethine imines in good to moderate yields upon isolation. The reaction conditions tolerated a variety of substituents (electron-withdrawing and electron-donating) being present on the aromatic ring (Scheme 3, examples 2a-2h). The C4-aminosubstituted pyrazolidin-3-one furnished the corresponding azomethine imine 2j in a reasonable 48% isolated yield. Structure variation in the benzylidene moiety can also be realized without major loss of reaction efficiency. The incorporation of the heteroaryl functionality such as pyridyl and furyl resulted in the formation of products 2k and 2l in reasonable yields. However, the alkylidene side chain proved to be more challenging. As demonstrated in Scheme 3 azomethine imine 2m was formed in situ with a good 70% yield and was found to be unstable during the isolation. Experiments of the photoinduced oxidative formation of azomethine imines 2 (Scheme 3) revealed that the major side product in the reaction is the formation of the corresponding aldehydes, caused by rather slower oxidation of the former azomethine imines 2 under optimized reaction conditions. Accordingly, we turned our attention to the development of a one-pot protocol involving dual catalysis, namely, eosin Y photocatalyzed formation of azomethine imines followed by Cu-catalyzed azomethine imine-alkyne cycloadditions. We anticipated that a fast click-reaction would prevent relatively slower overoxidation of 2 and therefore resulted in the formation of the corresponding cycloadducts 3 (Scheme 4). Applying the standard reaction conditions, eosin Y and TFA in acetonitrile in combination with Cu(0) proved to be the catalyst system of choice for formation of bicyclic product 3. In this manner we successfully developed a cascade protocol enabling the synthesis of an array of pyrazolo[1,2-a]pyrazoles 3 with good to excellent yields (Scheme 4). Experiments of the photoinduced oxidative formation of azomethine imines 2 (Scheme 3) revealed that the major side product in the reaction is the formation of the corresponding aldehydes, caused by rather slower oxidation of the former azomethine imines 2 under optimized reaction conditions. Accordingly, we turned our attention to the development of a one-pot protocol involving dual catalysis, namely, eosin Y photocatalyzed formation of azomethine imines followed by Cu-catalyzed azomethine imine-alkyne cycloadditions. We anticipated that a fast click-reaction would prevent relatively slower overoxidation of 2 and therefore resulted in the formation of the corresponding cycloadducts 3 (Scheme 4). Applying the standard reaction conditions, eosin Y and TFA in acetonitrile in combination with Cu(0) proved to be the catalyst system of choice for formation of bicyclic product 3. In this manner we successfully developed a cascade protocol enabling the synthesis of an array of pyrazolo[1,2-a]pyrazoles 3 with good to excellent yields (Scheme 4).
were proven to be more challenging, probably due to the low solubility of the starting material 1. Lower yields were obtained with the N1-alkyl, 5-methyl or phenyl substituted pyrazolidinones (Scheme 4, examples 3i-l). This could attribute to the partial endocyclic oxidation of the starting pyrazolidinones furnishing 1,2-dihydro-3H-pyrazol-3-ones 5. Thus, examining the crude reaction mixture of 3i revealed the endocyclic oxidation side product with m/z 189.1030 (see Supplementary Materials) indicating the competitive (exo)endocyclic oxidation pathways. Substituted azomethine imines 2 with variable substituents at the positions C1 (alkyl and aryl substituents), C4 (H or acylamino), and C5 (H, Me, i-Pr, or Ph) were in situ generated by photoinduced oxidation of the corresponding pyrazolidine-3-one 1. In situ generated azomethine imines were reacted with ynones under copper powder-catalyzed cycloaddition reaction conditions yielding the corresponding pyrazolo[1,2-a]pyrazoles 3. The relative configuration of compounds 3d, [1*] 3e, [2*] 3i-l is in agreement with the literature data and was established by NOESY NMR experiments. The workup was comprised from the removal of the heterogeneous Cu(0) catalyst by filtration followed by the evaporation of the solvent. Purification by flash chromatography (SiO 2 ) provided pure isolated products. Good yields of azabicycles 3 were achieved with both electron-withdrawing (Scheme 4, 3a,c,d) and electron-donating groups (Scheme 4, 3b) on the benzene ring of azomethine imines. The reaction also worked reasonably well in the case of compounds with unactivated alkylpyrazolidinones (Scheme 4, 3f-3h), however, longer side chains were proven to be more challenging, probably due to the low solubility of the starting material 1. Lower yields were obtained with the N1-alkyl, 5-methyl or phenyl substituted pyrazolidinones (Scheme 4, examples 3i-l). This could attribute to the partial endocyclic oxidation of the starting pyrazolidinones furnishing 1,2-dihydro-3H-pyrazol-3-ones 5. Thus, examining the crude reaction mixture of 3i revealed the endocyclic oxidation side product with m/z 189.1030 (see Supplementary Materials) indicating the competitive (exo)endocyclic oxidation pathways.
Following the investigation of the C5-monosubstituted pyrazolidinones in the one-pot protocol (Scheme 4, examples 3i-l), we subsequently examined the utility of our method by exploring 1-aryl-substituted pyrazolidinones 4 (Scheme 5), which could only undergo endocyclic oxidation delivering the corresponding pyrazolones 5. Reactions proceeded smoothly, delivering the desired products 5a-5c in 58-84% yields. Notably, electron-neutral and electron-deficient, 4a and 4b, substrates were found to be less reactive and needed longer reaction times (>24 h) compared to electron-rich 4c which was noticeably more reactive. C5-monosabstituted (methyl and phenyl) pyrazolidinones 4d and 4e were also suitable substrates but delivered the corresponding pyrazolones in rather lower yields. The formation of 5 is believed to proceed through the intermediate 4 which is generated via a single electron transfer (SET) event on N1 of pyrazolidinones 4. The formed intermediate is then isomerized into pyrazolone 5.
Catalysts 2020, 10, x FOR PEER REVIEW 6 of 10 products 5a-5c in 58-84% yields. Notably, electron-neutral and electron-deficient, 4a and 4b, substrates were found to be less reactive and needed longer reaction times (>24 h) compared to electron-rich 4c which was noticeably more reactive. C5-monosabstituted (methyl and phenyl) pyrazolidinones 4d and 4e were also suitable substrates but delivered the corresponding pyrazolones in rather lower yields. The formation of 5 is believed to proceed through the intermediate It is well known that the organic dye eosin Y can exist in four different forms in solutions depending on pH of the media, namely, in the spirocyclic form, the neutral form, the monoanionic, and the dianionic form. The spirocyclic form contains an interrupted conjugation of the fluorone moiety and thus would be photocatalytically inactive upon visible-light irradiation. The neutral form, on the other hand, exhibits only weak fluorescence when irradiated with visible light [45]. In our case, absorption properties of eosin Y in acidic medium, i.e., acetonitrile and TFA (0.5M solution) were examined. As evident from the UV-vis spectra (see Supplementary Materials) of eosin Y in in the mixture of MeCN/CF3CO2H (0.5M), the organic dye possesses significant absorption at 465-470 nm of the LED irradiation. Cyclic voltammetry of 1c with ferrocene standard shows E1/2 (1c (•+) /1c) = 0.73 V vs. SCE, which allows it to be oxidized by 3 EY* (E1/2 ( 3 EY*/EY (•-) ) = 0.83 V vs. SCE [46] (see Supplementary Materials).
When the reaction was performed under standard reaction conditions-(starting compound 1c (0.1 mmol), eosin Y-disodium salt (5 mol%), MeCN (1.0 mL), TFA (2.0 equiv.), air, 25-30 °C, 465-470 nm LED)-in the presence of a classical radical inhibitor (TEMPO), the formation of TEMPO-pyrazolone conjugate (m/z (MH + ); 349.2247 corresponding to C21H32ClN3O2) was detected by HRMS analysis in the crude reaction mixture. In absence of light irradiation (reaction running in the dark) only traces of product were observed, whereas in the absence of the photocatalyst no product was formed, which suggests the necessity of the photoredox catalyst and visible light for the reaction to occur. When the reaction was carried out under nitrogen atmosphere, only 5% of 2c was formed, proving the requirement of oxygen for the reaction to accrue. From the above controlled experiments and the literature reports [39,44,[47][48][49], the following plausible radical cation cascade reaction pathway has been suggested (Scheme 6). Initially, under visible-light irradiation, eosin Y (EY) is excited to 1 EY* which moves to its more stable 3 EY* through intersystem crossing (ISC) [50]. Subsequently, the excited state transforms pyrazolidinone 1c by SET into a nitrogen radical cation, with the concomitant reduction of 3 EY* in to the radical anion EY (•-) , which is afterwards oxidized by molecular oxygen regenerating the photocatalyst EY. On the other hand, deprotonation of the nitrogen radical cation generates the α-amino radical, that is further oxidized to the azomethine imine 2c by HO2 (•) species. The generated azomethine imine 2c subsequently enters copper catalyzed [3 + 2] cycloaddition reaction catalytic cycle furnishing pyrazolo[1,2-a]pyrazole 3a. It is well known that the organic dye eosin Y can exist in four different forms in solutions depending on pH of the media, namely, in the spirocyclic form, the neutral form, the monoanionic, and the dianionic form. The spirocyclic form contains an interrupted conjugation of the fluorone moiety and thus would be photocatalytically inactive upon visible-light irradiation. The neutral form, on the other hand, exhibits only weak fluorescence when irradiated with visible light [45]. In our case, absorption properties of eosin Y in acidic medium, i.e., acetonitrile and TFA (0.5M solution) were examined. As evident from the UV-vis spectra (see Supplementary Materials) of eosin Y in in the mixture of MeCN/CF 3 CO 2 H (0.5M), the organic dye possesses significant absorption at 465-470 nm of the LED irradiation. Cyclic voltammetry of 1c with ferrocene standard shows E 1/2 (1c (•+) /1c) = 0.73 V vs. SCE, which allows it to be oxidized by 3 EY* (E 1/2 ( 3 EY*/EY (•-) ) = 0.83 V vs. SCE [46] (see Supplementary Materials).
When the reaction was performed under standard reaction conditions-(starting compound 1c (0.1 mmol), eosin Y-disodium salt (5 mol%), MeCN (1.0 mL), TFA (2.0 equiv.), air, 25-30 • C, 465-470 nm LED)-in the presence of a classical radical inhibitor (TEMPO), the formation of TEMPO-pyrazolone conjugate (m/z (MH + ); 349.2247 corresponding to C 21 H 32 ClN 3 O 2 ) was detected by HRMS analysis in the crude reaction mixture. In absence of light irradiation (reaction running in the dark) only traces of product were observed, whereas in the absence of the photocatalyst no product was formed, which suggests the necessity of the photoredox catalyst and visible light for the reaction to occur. When the reaction was carried out under nitrogen atmosphere, only 5% of 2c was formed, proving the requirement of oxygen for the reaction to accrue. From the above controlled experiments and the literature reports [39,44,[47][48][49], the following plausible radical cation cascade reaction pathway has been suggested (Scheme 6). Initially, under visible-light irradiation, eosin Y (EY) is excited to 1 EY* which moves to its more stable 3 EY* through intersystem crossing (ISC) [50]. Subsequently, the excited state transforms pyrazolidinone 1c by SET into a nitrogen radical cation, with the concomitant reduction of 3 EY* in to the radical anion EY (•-) , which is afterwards oxidized by molecular oxygen regenerating the photocatalyst EY. On the other hand, deprotonation of the nitrogen radical cation generates the α-amino radical, that is further oxidized to the azomethine imine 2c by HO 2 (•) species. The generated is determined by addition of 1,3,5-trimethoxybenzene as standard. Compounds 2 were isolated by FC on silica with 10% solid NaHCO 3 (EtOAc, EtOAc/MeOH).

General Procedure for the Synthesis of pyrazolo[1,2-a]pyrazoles 3
In a 5-mL vial, starting compound (0.5 mmol) and eosin Y-disodium salt (5 mol%) are dissolved in MeCN (2.5 mL). TFA (2.0 equiv.), copper powder (<10 microns, 20 mg) and dipolarophile (1.5 eq.) are added and the solution irradiated with blue light under air on SynLED Photoreactor for 1-5 h until consumption of starting material is complete according to TLC. K 2 HPO 4 (2 equiv.) is added and the mixture stirred in the dark for an additional 16 h. Products 3 are isolated by CC (EtOAc/hexanes).

General Procedure for the Synthesis of pyrazolones 5
In a 5-mL vial, starting compound (0.5 mmol) and eosin Y-disodium salt (2 mol%) are dissolved in MeCN (2.5 mL). The solution is irradiated with blue light under air on SynLED Photoreactor overnight (18-24 h). After evaporation of the solvent, the NMR yield is determined by addition of 1,3,5-trimethoxybenzene as standard. Products 5 can be isolated by FC (EtOAc/hexanes).

Conclusions
In summary, an efficient and mild visible-light-mediated aerobic transformation of pyrazolidine-3-one derivatives is disclosed. The methodology enables the generation of azomethine imines as versatile intermediates. The photoredox cycle is merged with copper powder catalyzed [3 + 2] cycloaddition reaction yielding the corresponding cycloadducts in a one-pot synthetic protocol. The complexity of the product delivered under this mild reaction conditions makes these new methods an attractive alternative for the formation of functionalized pyrazolo[1,2-a]pyrazoles. The methodology was also applied to 1-aryl-substituted pyrazolidinones yielding pyrazolones in yields up to 84%.