Squaramide‐IRMOF‐16 Analogue for Catalysis of Solvent‐Free, Epoxide Ring‐Opening Tandem and Multicomponent Reactions

Tandem and multicomponent one‐pot reactions are highly attractive because they enable synthesis of target molecules in a single reaction vessel. However, they are difficult to control, as they can lead to the formation of many undesired side‐products. Herein we report the use of metal‐organic framework (MOF) pores decorated with organocatalytic squaramide moieties to confine ring‐opening epoxide reactions of diverse substrates. Controlled mono‐addition or tandem reactions inside the pores yield 1,2‐aminoalcohols or 1,2,2′‐aminodialcohols, respectively, in good yields. In addition, this squaramide‐functionalised MOF enables catalysis of higher‐complexity multicomponent reactions such as the catalytic ring‐opening of two different epoxides by a single amine to afford 1,2,2′‐aminodialcohols.


Squaramide-IRMOF-16 Analogue for Catalysis of Solvent-Free, Epoxide Ring-Opening Tandem and Multicomponent Reactions
Claudia Vignatti + , [a] Javier Luis-Barrera + , [b] Vincent Guillerm, [a] Inhar Imaz,* [a] Rubén Mas-Ballesté,* [c, d] José Alemán,* [b, d] and Daniel Maspoch* [a, e] Tandem and multicomponent one-pot reactions are highly attractive because they enable synthesis of target molecules in a single reaction vessel. However, they are difficult to control, as they can lead to the formation of many undesired sideproducts. Herein we report the use of metal-organic framework (MOF) pores decorated with organocatalytic squaramide moieties to confine ring-opening epoxide reactions of diverse substrates. Controlled mono-addition or tandem reactions inside the pores yield 1,2-aminoalcohols or 1,2,2'-aminodialcohols, respectively, in good yields. In addition, this squaramidefunctionalised MOF enables catalysis of higher-complexity multicomponent reactions such as the catalytic ring-opening of two different epoxides by a single amine to afford 1,2,2'-aminodialcohols.
Tandem reactions are among the best strategies to achieve molecular complexity in a single process. [1] They comprise two or more consecutive independent reactions, which are catalysed by one or more catalysts. Each catalyst produces an intermediate that is further transformed by a second catalytic cycle to give the final product. This translates to lower requirements for solvent, time and energy and to less waste relative to traditional processes. Consequently, tandem reac-tions have peaked the interest of numerous industries, [2] especially in their solvent-free form. [3] 1,2-amino alcohols (3) and 1,2,2'-aminodialcohols (4) are structural subunits that are widespread in natural products of industrial relevance. [4] Some of these natural products include (S,R,R,R)-Nebivolol, which is a b 1 -adrenergic receptor blocker; [4e] Bestatin, which is an aminopeptidase inhibitor that exhibits immunomodulatory activity; [4d] Sphingosine, which is a a class of cell membrane lipids; [4d] and Cytoxazone, which is an immunomodulatory. [4d] They are also important synthetic intermediates for biologically active compounds, [4] stationary phases in HPLC, [5] and chiral ligands (e. g. Oxazaborolidine derivatives) [6a] or auxiliaries in asymmetric reactions. [6] 1,2-aminoalcohols and 1,2,2'-aminodialcohols can each be readily prepared via ring-opening of epoxides by amines. However, controlling the reaction of the amine (i. e. mono-vs. di-addition) to the epoxide is difficult, leading to mixtures of the two types of compounds.
Herein we show that confining the aforementioned reaction to metal-organic framework (MOF) [7] pores decorated with organocatalytic squaramide moieties enables control over the formation of the mono-or di-addition products (see below). Furthermore, it also allows for the selective synthesis of heterogeneous double-addition products via multicomponent reactions in which two different epoxides are opened by a single amine (see below). Recently, Hupp, Farha, Mirkin et al. [8] and Cohen et al. [9] demonstrated that squaramide moieties can be incorporated into MOFs by post-synthetic modification of UiO-67 and by using a tetracarboxylate squaramide-based linker to produce a new Cu(II)-based MOF showing a pore diameter of~8 Å 8 Å . Both squaramide-functionalised MOFs [10] were successfully tested as catalysts for Friedel-Crafts reactions between indoles and b-nitroalkenes. For our targeted catalytic reactions, we constructed a squaramide-functionalised IRMOF-16 analogue (hereafter called Sq_IRMOF-16) because it shows a three-dimensional mesopore system in which the squaramide moieties are totally accessible in all three dimensions and are well separated to avoid any self-quenching phenomena. In addition, the pore diameter is~17 Å 17 Å , which is sufficiently large to host the intermediates produced during the tandem reactions. The linker (3,4-dioxocyclobut-1-ene-1,2-diyl)bis(azanedyil)-p-dibenzoic acid, hereafter called L1) was designed to resemble, both in topology and in length, to the p,p'-terphenyl dicarboxylic acid (tpdc), which is the linker used to synthesise IRMOF-16 ( Figure 1b). [11] Moreover, and in contrast to a previously reported linker [9] in which the squaramide moiety is in the meta position to the acid, in L1 the squaramide moiety is  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57 para to the carboxylic group. Therefore, our design leads to a more acidic NH proton than in the meta case. We began by synthesising L1, using slight modifications of previously reported procedures. [12] Then, Sq_IRMOF-16 was synthesised by heating a mixture of L1 and Zn(NO 3 ) 2 in N,Ndimethylformamide (DMF) at 85 8C for 7 days. After this period, yellow cubic crystals of Sq_IRMOF-16 were harvested (yield = 53 %). As expected, the experimental powder X-ray diffraction (PXRD) pattern of Sq_IRMOF-16 was in excellent agreement with the one calculated from the envisioned squaramide-based IRMOF-16 (Figure 1c, see also Supporting Information, Figure S1). The squaramide-based IRMOF-16 model was constructed from the experimental IRMOF-16 structure [11] by ligand replacement, respecting the symmetry of IRMOF-16 (Pm-3m space group). This step was followed by a molecular mechanics energy minimisation to improve the geometry of the bonds within the framework using the Forcite tool of the Materials Studio software (Biovia). [13] Therefore, analogously to IRMOF-16, Sq_IRMOF-16 comprises a zinc-metal cluster (Zn 4 O) bridged by six dicarboxylate linkers that form a network with pcu topology. The network is a three-dimensional mesopore system (pore size:~17 Å 17 Å ) in which the squaramide moieties point towards the pores and therefore, are totally accessible in all three dimensions (Figure 1b).
For the catalytic experiments, we carefully dried Sq_IRMOF-16 dried under inert atmosphere and then, immediately mixed it with the other reagents (Supporting Information, Figure S2). It is worth to mention that this drying step was critical, as Sq_ IRMOF-16 tends to become amorphous upon exposure to vacuum, and to transform into an unknown crystalline phase upon contact with water (Supporting Information, Figure S3). In order to verify that Sq_IRMOF-16 remained stable during the catalytic processes, it was recovered from the reaction media after the catalytic runs and its crystalline phase was confirmed by XRPD (Supporting Information, Figure S4). Additional experiments proved that the catalytic activity of Sq_IRMOF-16 was not related to the degradation or leaching of molecular species under the reaction conditions. [14] As a first approach to studying the catalytic behaviour of Sq_IRMOF-16, we monitored the kinetics of the reactions of each amine (1 a, R=Me, or 1 b, R = t-Bu) with each epoxide (2 a, R = Et or 2 b, R = C 10 H 21 ) at 60 8C (Figure 2b). We introduced to the reaction medium a 5 mol % content of catalytic centers, which are included in the structure of Sq_IRMOF-16; that is, 2.9 mg of Sq_IRMOF-16 that corresponds to 0.005 mmol of catalytic units were used to catalyze the reaction of 0.1 mmol of the corresponding aniline with an excess of epoxide. Figure 2c is a plot of the kinetics for each mono-addition product, which was the major species at 8 hours of reaction. Here, the   1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57 performance of Sq_IRMOF-16 was also compared with the molecular squaramide C as catalyst (Figure 2a). Using both catalysts, we studied the reaction of 1 a with 2 a (compare the blue dashed line with the blue solid one) and 2 b (compare the red dashed line with the red solid one). The reactions barely progressed when using C, probably due to the auto-selfaggregation and poor solubility of the catalyst. In quite contrast, the use of Sq_IRMOF-16 enhanced both kinetics and yields of these reactions. Moreover, when using Sq_IRMOF-16, we observed that epoxide 2 a (R 2 = Et) appeared to react better than epoxide 2 b (R 2 = C 10 H 21 ), as observed in Figure 2c (compare the solid blue line with the solid red one, or the solid grey line with the solid orange one). Likewise, amine 1 a (R 1 = Me) typically reacted faster than amine 1 b (R 1 = t-Bu), also evidenced in Figure 2c (compare the solid blue line to the solid grey one, or the solid red line to the solid orange one. Interestingly, in the case of the use of the smaller epoxide 2 a in their reaction with 1 a and 1 b (blue and greys lines), we also found a significant amount of the dialkylated products 4 a and 4 b (see Supporting Information). Altogether, these observations suggest that there is a size discrimination effect when Sq_ IRMOF-16 is used, which is probably due to the lower diffusion rates of the bulkier substrates. These differences confirm that the catalytic processes occur inside the pores of Sq_IRMOF-16 rather than on its external crystal surfaces.
Interestingly, we observed that once the mono-addition products 3 were obtained, the bis-addition products, homodisubstituted amino diols 4, began to form. Figure 3 shows a series of tandem reactions of the amines 1 a-b and epoxides 2 a-f to form the diols 4 a-g catalysed by Sq_IRMOF-16 (5 mol %) to test its catalytic utility. Remarkably, this reaction tolerated many combinations of reagents. The times required to obtain optimised yields of a series of diols 4 correlated to the size (4 a-4 d; Figure 3, top row) and/or polarity (4 e-4 g; Figure 3, bottom row) of the substrates. For example, comparing the synthesis of 4 a with that of 4 b reveals that ethylepoxide (2 a) reacted faster with para-methoxy aniline (1 a) than with para-tert-butyl aniline (1 b). Similarly, to 4 b, the diols 4 c (from 1 a and 2 b) and 4 d (from 1 b and 2 c) required 3 days and 4 days, respectively, to reach moderate yields. We attributed these low reaction rates and moderate yields to the steric bulk and hydrophobicity of the alkyl chains in epoxides 2 b (R 2 = C 11 H 23 ) and 2 c (R 2 = C 10 H 21 ), which could hamper the diffusion of each epoxide through the pores of Sq_IRMOF-16.
In the above reactions, we also found that the bulkier epoxide 2 e reacted at a similar reaction rate than did the smaller epoxide 2 a. We ascribed this fact to the greater polarity of the -CH 2 OTBDMS group in 2 e relative to the -Et group in 2 a, which may help the diffusion of 2 e through the pores of Sq_ IRMOF-16. Consistent with our hypothesis, the more polar epoxides 2 f (R 2 = CH 2 HNBoc) and 2 g (R 2 = CH 2 CO 2 Et) gave neartotal conversion (yields > 90 %) to their corresponding diols 4 f and 4 g, respectively, after only 22 h. We next evaluated the capacity of Sq_IRMOF-16 to catalyse multicomponent reactions of higher complexity. To this end, we used three reagents (one amine reacted sequentially with two epoxides) to generate heterogeneous diols in one-pot multicomponent reactions. This approach typically requires less energy and generate less waste than step-reactions which needs multiple purification processes. However, a drawback of one-pot reactions for heterogeneous additions is that they demand strict control of the chemistry. In our case, to avoid the formation of undesired side-products, a single mono-addition intermediate 3 had to be generated first. Once 3 had been formed in the reaction media, via one pot process (i. e. without any purification), other epoxides can be added to obtain the desired hetero-disubstituted amino diols 5 ( Figure 4).
In conclusion, we have synthesised a squaramide-functionalised IRMOF-16 analogue, Sq_IRMOF-16, for use as a catalyst in the ring-opening of epoxides by nucleophilic amines. Sq_ IRMOF-16 does not undergo the self-aggregation phenomena usually observed for squaramides in solution; in fact, this heterogeneous catalyst is superior to its molecular squaramide analogue. The pores in Sq_IRMOF-16 are sufficiently large to catalyse the ring-opening of diverse epoxides using different amines. We have demonstrated the catalytic activity of Sq_ IRMOF-16 in the synthesis of simple, tandem and multicomponent epoxide ring-openings under solvent-free conditions and in good yields. The evidences suggest that these reactions are confined to the squaramide-functionalised pores, as Sq_IRMOF-16 shows size-and polarity-discrimination effects. Given that many organocatalytic moieties can be introduced into MOF pores, we are confident that MOF-based catalysts such as Sq_IRMOF-16 should help to expand the scope of heterogeneous catalysis in one-pot reactions.