Efficient NiII 2 LnIII 2 electrocyclization catalysts for the synthesis of trans-4 , 5-diaminocyclopent-2-enones from 2-furaldehyde and primary or secondary amines

A series of heterometallic coordination clusters (CCs) [Ni2Ln III 2(L1)4Cl2(CH3CN)2] 2CH3CN [Ln = Y (1Y), Sm (1Sm), Eu (1Eu), Gd (1Gd), or Tb (1Tb)] were synthesized by the reaction of (E)-2-(2-hydroxy-3-methoxybenzylidene-amino)phenol) (H2L1) with NiCl2·6(H2O) and LnCl3·x(H2O) in the presence of Et3N at room temperature. These air-stable CCs can be obtained in very high yields from commercially available materials and are efficient catalysts for the room-temperature domino ring-opening electrocyclization synthesis of trans-4,5-diaminocyclopent-2-enones from 2-furaldehyde and primary or secondary amines under a non-inert atmosphere. Structural modification of the catalyst to achieve immobilization or photosensitivity is possible without deterioration in catalytic activity. ■ INTRODUCTION Polynuclear coordination clusters (CCs), assembled from organic ligands, transition-metal elements (3d), and/or lanthanide (4f) elements, are a burgeoning class of functional materials with fascinating structures. These materials are already used in areas ranging from molecular magnetism, biology, molecular imaging, magnetic resonance, and luminescence. However, one of the most common applications for coordination compounds, catalysis, is much less studied for polynuclear 3d/4f heterometallic CCs with classical N,O-donor ligands than for organometallic compounds. The difference in the growth rates between the two research fields may reflect the fact that organometallic compounds are mainly of low nuclearity, so it is easier to determine a possible mechanism for the catalytic procedure. Electrocyclizations are one of the most valuable transformations for the construction of carbon−carbon bonds. The Nazarov cyclization exemplifies such an electrocyclization, providing access to important and useful cyclopentones, and Lewis acids catalyze this cyclization. The interest in bifunctionalized cyclopentenones arises from the fact that they are versatile building blocks in the synthesis of natural products such as (+)-Nakadomarin A. Batey and Li used a variety of Lewis acids for the synthesis of trans-4,5-diaminocyclopentanones (Scheme 1) from the domino condensation/ringopening/electrocyclization reaction of furfuraldehyde with amines. In this study, Dy(OTf)3 and Sc(OTf)3 were found to be the best catalysts, with 10% loadings. However, Sc(OTf)3 gave higher yields with primary anilines (R = H; see Scheme 1) than Dy(OTf)3. This reaction proceeds in high yields with the use of acidic ionic liquids under solvent-free conditions or with the use of N-sulfonylimine, with p-toluenesulfonamide acting as a leaving group. However, in both cases, the reaction works only for secondary amines. We have recently developed a new generation of 3d/4f CCs and have explored their catalytic behavior. The Schiff base H2L1 44 (Scheme S1 in the Supporting Information), along with 3d and 4f ions affords tetranuclear M2Ln III 2 (M = Co or Ni) CCs possessing a defect dicubane topology with two divalent 3d ions (center) and two trivalent 4f ions (wings) (Scheme S1, middle). In this specific topology, five out of six, and six out of seven or eight, coordination sites are occupied by H2L1 for the 3d and the Dy centers, respectively (Scheme S1, right). Recently, we reported that compound [Ni2Dy III 2(L1)4Received: March 22, 2016 Published: June 29, 2016 Scheme 1. Schematic Representation of trans-4,5Diaminocyclopentanones Article


Introduction
Polynuclear Coordination Clusters (CCs), 1,2 assembled from organic ligands, transition metal (3d) and/or lanthanide (4f) elements, are a burgeoning class of functional materials with fascinating structures. 3-8 These materials are already used in areas ranging from molecular magnetism, 9-17 biology, [18][19][20] molecular imaging, 21 magnetic resonance 22 and luminescence. [23][24][25][26][27] However, one of the most common applications for coordination compounds, catalysis, is much less studied for polynuclear 3d/4f heterometallic CCs with classical N,O-donor ligands, 28-34 than for to organometallic compounds. 35,36 The difference in the growth rates between the two research fields may reflect the fact that organometallic compounds are mainly of low nuclearity and so it is easier to determine a possible mechanism for the catalytic procedure.
Electrocyclizations are one of the most valuable transformations for the construction of carbon carbon bonds. The Nazarov cyclization exemplifies such an electrocyclization providing access to important and useful cyclopentones and Lewis acids catalyze this cyclization. [37][38][39] The interest in bifunctionalized cyclopentenones arises from the fact that they are versatile building blocks in the synthesis of natural products such as (+)-Nakadomarin A. 40 Batey and Li used a variety of Lewis acids for the synthesis of trans-4,5-diaminocyclopentanones (Scheme 1) from the domino condensation/ring-opening/electrocyclization reaction of furfuraldehyde with amines. 41 In this study, Dy(OTf)3 and Sc(OTf)3 were found to be the best catalysts, with 10% loadings. However Sc(OTf)3 gave higher yields with primary anilines (R 2 =H, Scheme 1) than Dy(OTf)3. This reaction proceeds in high yields with the use of acidic ionic liquids under solvent-free conditions 42 or with the use of N-sulfonylimine, with p-toluenesulfonamide acting as a leaving group. 43  loading. This CC, unlike Dy(OTf)3, could catalyze the reaction with secondary and primary amines. 41,47 In our quest to develop these catalytic species further and to a) seek lower cost materials and b) investigate its possibility of immobilization, we decided to i) extend our study to other lanthanides and ii) to tune the organic periphery of the H2L1 ligand. Synthetic, catalytic and theoretical issues are discussed in this report.

Experimental Section
Materials and methods

Results and Discussion
With the goal of developing effective molecular catalysts with low cost, we reasoned that replacement of Dy III with lanthanides of lower cost would allow us to evaluate the relationship between cost and effectiveness. Thus, we attempted to isolate and characterize the analogues of compound 1Dy with Y (1Y), Sm (1Sm), Eu (1Eu), Gd (1Gd), and Tb (1Tb) following the similar reaction procedure that gave 1Dy. 47 The synthetic protocol specifies that when the blend Dy(OTf)3 / Ni(ClO4)2·6H2O or NiCl2·6H2O / H2L1 / Et3N (molecular ratio 1 / 1 / 2 / 5) in CH3CN (20 mL) is refluxed for one hour results, after filtration, in yellow-greenish crystals in moderate yield. As the reactions to produce the 1Y, 1Sm, 1Eu, 1Gd, and 1Tb  In our attempts to grow crystals, several unexpected products were structurally characterized, indicating that the synthesis of these species is not a simple task. 69  The next step was to characterize the catalytic activity of these species using the catalytic protocol developed in our previous study. 47 In order to ascertain that the catalytic activity of these species is driven solely by the 4f ions, various blank tests were performed (Table S1). First, three Ni II salts with three different loadings (Table S1, Entries 2 -10) were used in the prototype reaction.
NiCl2 was found to catalyze the reaction with very high loadings and very low yields after 24 hours (Table S2, Entry 2) reflecting that its contribution is negligible for the shorter time period. An in situ mixture of Ni salt and ligand (Table S2,  gave none of the anticipated product after 24h. In Batey's protocol, 41 as Dy(OTf)3 was found to be an excellent catalyst using a loading of 10 mol%, our next step was to employ other lanthanide salts (  1, entry 7). The higher efficacy of Y III than Ln III in Lewis acid catalysis, has been previously reported, 71,72 but no rationale has been proposed. All attempts to recover the 1Ln catalyst in order to reuse it, were unsuccessful.  The excellent results using even lower catalyst loadings (0.5% for 1Y instead of 1.0% for 1Dy) and the cheaper lanthanide source, led us to consider the possibility that 1Y could be structural modified to achieve immobilization and demonstrate photosensitivity. We recently reported a library of nineteen modified Schiff-base organic ligands that offer coordination environment similar to that in H2L1. 69 Among these, the three organic ligands shown in Scheme 2 were selected for the synthesis of the corresponding tetranuclear CCs. Other considerations were i) to increase the solubility as well to offer the possibility for deposition of these species on surfaces by the introduction of an allyl group (H2L2) and ii) to create "photosensitive" catalysts by inclusion of a naphtho group (H2L3) and iii) to combine both these features (H2L4).  (Figure 1C), respectively. These compounds were characterized by X-Ray crystallography (Figure 1)  The catalytic activity of 2Y, 3Y and 4Y towards the prototype reaction is shown in Table 2.
Compounds 2Y and 4Y are not so efficient as to 1Y (1.0% instead of 0.5% loading and slightly lower yields), but these results can be attributed to the presence of the sterically demanding naphtho groups. When 3Y was employed as catalyst (Table 2) yields similar to those with 1Y were obtained. As we reported before, 47 when the reaction takes place with primary amines the corresponding deprotonated Stenhouse salts are formed (Entries 6-9). The ring-closing of the salt is promoted by the use of very dilute HCl leading to the corresponding trans-4,5diaminocyclopent-2-enones.
To explore further the possible immobilization of these catalysts, we attempted to synthesize the novel organic ligands H4L5 and H4L6 shown in Scheme 3 (see ESI for synthetic details). Ligand H4L5 offers a coordination environment similar to that in H2L1 and thus it was expected to form coordination polymers upon complexation with the Ni II and Y III salts. However, H4L5 is only slightly soluble in polar solvents such as N,N'-DMF or DMSO. Organic ligand H4L6 was synthesized expecting a higher solubility, but like H4L5 it is only slightly soluble in polar solvents.
Our efforts to synthesize the corresponding coordination polymers have not yet been successful.  and H2L4 (lower) and bound (left) and its free conformation (right).

Conclusions
The collective data on this simple and high yielding isoskeletal series of Ni II 2Ln III 2 CCs, supported by structurally homologous Schiff base ligands, shows remarkable catalytic activity in the formation of trans-4,5-diaminocyclopent-2-enones from 2-furaldehyde and primary or secondary amines. The Ni II contribution is negligible and therefore it is safe to deduce that the domino reaction is driven solely by the 4f or Y ion. Remarkably, the employment of Y III instead of Dy III makes possible a lower loading and lower cost catalyst. We show that the functionalization of the organic skeleton is feasible (2Y, 3Y and 4Y), without impairing the catalytic activity.
Moreover, our results suggest that the modified species 2Y and 4Y, that bear allyl groups, can be immobilized by deposition on surfaces. Further work on this aspect is in progress. We are currently exploring the possibilities of utilizing 3d/4f or 3d/Y CCs as tandem catalysts in other organic transformations.
Supporting Information.