Spiroacetals appear in a wide range of natural products and biologically active molecules . Well-known examples include the marine toxins okadaic acid , isolated from the sponge Halichondria okadai, and azaspiracid-1 , obtained from blue mussels (Mytilus edulis), as well as hydantocidin , found in Streptomyces hygroscopicus (Fig. 1). Besides their intriguing biological activities, these molecules are an amazing example for structural complexity created by nature. Most common are [O,O]-spiroacetals featuring tetrahydrofuran or tetrahydropyran rings, but unsaturated systems, as well as, [N,O]-spiroacetals are also observed. Occasionally, the spiroacetal motif contains even more than two heteroatoms.
As a consequence of the interesting biological and structural properties of spiroacetals, there is a high demand for efficient methods to synthesize these privileged scaffolds. Besides annulation of heterocyclic substrates, the two-fold cyclization of ketones or alkynes bearing two nucleophilic sidechains is of particular importance [5–9]. For the spiroacetalization of acetylenic diols and related substrates, the triple bond has to be activated for nucleophilic attack with a suitable transition metal catalyst. Already in 1983, Utimoto reported the palladium-catalyzed spirocyclization of diol 1 which afforded spiroacetal 2 with high yield (Scheme 1) . Since the beginning of this millennium, it was established that gold catalysts are highly suitable for the activation of multiple bonds (in particular alkynes) [11–17]. Not surprisingly, this method was also applied to the spiroacetalization of acetylenic diols; in a seminal contribution, Liu and De Brabander obtained a mixture of spiroacetals 4 and 5 from diol 3 in the presence of a cationic gold catalyst . Several applications of the method (also in natural product synthesis) were reported in recent years [19–25] and the highest turnover numbers published for homogeneous gold catalysts have been achieved for such spiroketalizations [26, 27]. In contrast to this, gold-catalyzed approaches to [N,O]-spiroacetals are rare [28, 29].
Based on 15 years’ experience in the gold-catalyzed activation of unsaturated substrates (mostly allenes) for regio- and stereoselective cyclization reactions [30–33], we have recently expanded our synthetic repertoire to the spiroacetalization of functionalized alkynes. Main objectives of these investigations are the application of the method to new types of [O,O]- and [N,O]-spiroacetals which are of interest as molecular scaffolds in medicinal chemistry, as well as, the use of sustainable reaction conditions by employing recyclable gold catalysts in water as bulk reaction medium.
Results and discussion
As a starting point in our synthetic approach towards spiroacetals, we briefly studied the cyclization of hept-3-yn-1,7-diols, as well as, hept-2-en-4-yn-1,7-diols, to [O,O]-spiroacetals 7 bearing two five-membered rings (Scheme 2) .
In analogy to the corresponding intermolecular acetalization of homopropargylic alcohols , the gold-catalyzed formation of spiroacetals 7 with two dihydro- or tetrahydrofuran rings is facile. The best results were obtained with the cationic gold catalyst formed in situ from Ph3PAuCl and AgOTf. In diethyl ether as solvent, good yields of the desired 1,6-dioxaspiro[4.4]nonanes 7 were obtained within 5 min at room temperature. Diastereoselectivities, however, were low and did not exceed a ratio of 2:1. It must be noted that in the case of hept-2-en-4-yn-1,7-diols, the presence of two substituents in position 1 is required for the formation of unsaturated spiroacetals 7e–g; in the absence of these substituents, cycloisomerization of the 2-en-4-yn-1-ol fragment affords the corresponding furans [13, 36].
Next, we extended the method to the spiroacetalization of amino-substituted alkynols 8 which afford the desired [N,O]-spiroacetals 9 bearing five- or six-membered rings with moderate to high yield (Scheme 3) . Due to a strong deactivation of the gold catalyst, no reaction occurs in the absence of a protecting group at nitrogen . An interesting dependence of the spirocyclization on the protecting group was observed: whereas the Boc-protected spiroacetal 9a was isolated with high yield, but low diastereoselectivity, the opposite result was obtained for the tosylated counterpart 9b. Also, it should be noted that the reactivity of six-membered ring formation is rather low, resulting in moderate yields of spiroacetals 9e and 9f. In general, diastereomeric ratios of the [N,O]-spiroacetals 9 are considerably higher than those of the corresponding [O,O]-spiroacetals 7. In the case of the products 9b and 9e, the anti-configuration of the major diastereomer was proven by X-ray crystallography.
We then turned our attention to the gold-catalyzed synthesis of spiroacetals with three heteroatoms. Formal replacement of the N-hydrogen atom of alkynes 8 by a hydroxy group leads to alkynols of the type 10 with a hydroxylamine sidechain. These can be synthesized by addition of zinc acetylides to protected nitrones . Exposure of 10 to different gold catalysts revealed a low reactivity; the best results were obtained with phosphite–gold(I) complex A in the presence of AgSbF6 which afforded the spirocyclic isoxazolidine 11 with 93% yield after 2 days at room temperature (Scheme 4) . Increasing the temperature to 50°C shortened the reaction time to 4 h and also led to an improved diastereoselectivity of 76:24.
In order to improve the sustainability of chemical processes, it is highly desirable to perform not just one, but several transformations in one flask. This procedure saves time and reduces waste as isolation and purification of intermediates is no longer necessary. Indeed, the gold-catalyzed synthesis of spirocyclic isoxazolidine 11 can be performed as a two-component reaction by starting with pent-4-yn-1-ol 12 and nitrone 13 (Scheme 5) . Strong heating in toluene under microwave conditions, and the presence of an additional Lewis acid to promote the acetylide addition to the nitrone, are required. Among several Lewis acids studied, indium trichloride gave the best result and afforded spiroacetal 11 with 60% yield. Remarkably, it is even possible to perform the nitrone formation from isobutyraldehyde 14 and PMB-protected hydroxylamine 15 under these conditions. Thus, the three-component coupling of 12, 14, and 15 in the presence of the cationic gold catalyst and InCl3 gave spirocyclic isoxazolidine 11 with 43% yield and a diastereomeric ratio of 81:19. It should be noted that under these conditions four bonds are formed with high efficiency in one pot.
In 2012, Ohno and coworkers have developed an efficient and highly sustainable approach to dihydropyrazoles by gold-catalyzed three-component annulation of alkynes with hydrazines and aldehydes or ketones . This method was applied to the one-pot synthesis of different indazole and indole derivatives . In a cooperation project, we have recently extended this method to a three-component spirocyclization of alkynols 16, aldehydes 17, and hydrazine derivatives 18 which readily affords hitherto unknown spirocyclic pyrazolidines 19 (Scheme 6) . In contrast to the synthesis of [O,O]-spiroacetals 11, moderate heating of the three components in the presence of phosphite–gold complex A and silver hexafluoroantimonate is sufficient to afford the [N,O]-spiroacetals 19 in up to 97% yield.
The spirocyclic pyrazolidines 19 were obtained with diastereomeric ratios of up to 4:1. The relative configuration of the major diastereomer of 19d was determined by X-ray crystallography to be (3RS,5SR). The scope of the three-component spiroacetalization is excellent and allows introduction of substituents in most positions of the spirocycles 19. This renders the method highly valuable for applications in combinatorial or medicinal chemistry. Thus, a wide variety of aliphatic (19a/e), aromatic (19b/c/d/f), and heteroaromatic aldehydes (19g/h) is tolerated. Notably, fluorinated aryl groups (19c), as well as, bromide (19d) can be introduced without difficulty, the latter offering a handle for further functionalization. Structural variations of the alkynol are possible as well and include the introduction of substituents at different positions and extension of the tether between triple bond and hydroxy group. The hydrazine has to bear an electron-rich and an electron-deficient group. The former can be benzyl or p-methoxybenzyl; for the latter, various carbamates can be employed. This opens up different options for further transformation of the spirocycles. For example, hydrogenative debenzylation of 19a furnished the monoprotected pyrazolidine 20 with almost quantitative yield (Scheme 7). In contrast, removal of the Boc group under acidic conditions led to a mixture containing 50% of the ring-opened product 21. Obviously, the presence of a protecting group at the hemiaminal nitrogen is important for the stability of the spirocyclic pyrazolidine .
From the mechanistic point of view, it is assumed that the transformation starts with the known gold-catalyzed cycloisomerization of alkynol B to enol ether E via intermediates C and D (Scheme 8) [35, 42, 43]. Here, alcohols B/C and/or residual water serve as proton shuttle . A subsequent [3+2]-cycloaddition with azomethine ylide F may follow a stepwise (via intermediate G) or concerted pathway (via transition state H) . This mechanistic model is based on NMR-spectroscopic investigations which revealed a rapid consumption of the alkynol whereas the hydrazine is consumed at a slower rate. Moreover, an intermediate was observed in the 1H-NMR at δ~3.5 which may be attributed to an enol ether. Unfortunately, attempts to perform the [3+2]-cycloaddition with preformed enol ethers have failed due to the known instability of these substrates .
So far, we have performed the spiroacetalizations by a classical approach using well-established gold catalysts in organic solvents. Under these traditional conditions, however, it is impossible to recycle the gold catalyst as it is rapidly reduced to catalytically inactive metallic gold [15–17, 46]. In recent years, we have developed several approaches to solve this fundamental limitation of homogeneous gold catalysis. These involve the use of ionic liquids [47, 48] or micelles [49–51] as reaction medium, as well as, the design of recyclable water-soluble gold-NHC catalysts [52, 53]. Gratifyingly, these methods can also be applied to the synthesis of spiroacetals. For example, the cycloisomerization of alkynediol 6a to [O,O]-spiroacetal 7a can be carried out in water in the presence of the ammonium salt-tagged gold-NHC complex J (Scheme 9) . After extraction of the product, the gold catalyst solution can be reused at least five times without loss of activity. It should be noted that an activation the complex J with a silver salt under formation of a cationic gold species is not necessary.
In recent years, Lipshutz and coworkers have pointed out nanomicelles (formed in bulk water from vitamin E-based amphiphiles) as adequate reaction medium for various transition metal-catalyzed transformations . The surfactant provides a very limited amount of an organic medium in which the reactions can take place under high internal concentrations and mild conditions. In the case of gold-catalyzed transformations, micellar catalysis enables an efficient stabilization of the catalyst which thereby becomes recyclable [49, 55]. Dehydrative reactions are particularly interesting applications of micellar gold catalysis. In organic media, dehydrative reactions are typically driven by the presence of a dehydrating agent, such as molecular sieves. In contrast, the hydrophobic effect that exists within the lipophilic cores of nanomicelles causes the water formed in a dehydrative reaction to be expelled, thereby driving the transformation to completion. We have applied this principle to the dehydrative spirocyclization of acetylenic triols which affords unsaturated [O,O]-spiroacetals and one equivalent of water . For example, in the presence of Ph3PAuCl/AgOTf and D-α-tocopherol-polyethyleneglycol-750-succinate monomethyl ether (TPGS-750-M) as amphiphile, triol 22 is smoothly cyclized to spiroacetal 23 within 5 h at room temperature in water as bulk (external) medium (Scheme 10) .
Even highly demanding reactions such as the three-component coupling affording spirocyclic pyrazolidines 19 (Scheme 6) can be carried out under the challenging conditions of micellar catalysis. As a proof of principle, we could demonstrate that the reaction of isobutyraldehyde, pent-4-yn-1-ol, and benzyl/Cbz-protected hydrazine with cationic gold catalyst K in an aqueous medium containing 5% polyoxyethanyl α-tocopheryl sebacate (PTS) and 3 M NaCl afforded spiroacetal 24 with 35% yield after 20 h at 50°C (Scheme 11) . The presence of salt serves to improve the reactivity of the micellar catalyst system.
The gold-catalyzed spiroacetalization of suitable functionalized alkynes opens a versatile and efficient access to different types of [O,O]- and [N,O]-spiroacetals. Simple spiroacetals 7 and 9 with two heteroatoms are readily obtained by regioselective cyclization of acetylenic diols 6 and aminoalcohols 8, respectively. A particularly effective synthetic approach to the hitherto unknown spiroacetals 11 and 19 bearing three heteroatoms was realized by gold-catalyzed three-component coupling of alkynols, aldehydes, and protected hydroxylamine or hydrazine derivatives. The sustainability of these spirocyclizations can be improved even further by using recyclable gold catalysts in water or nanomicelles as reaction medium. Further work devoted to an improved substrate scope, reactivity, sustainability, and stereoselectivity of gold-catalyzed spirocyclizations is in progress.
A. S. K. Hashmi, F. D. Toste, eds. Modern Gold Catalyzed Synthesis, Wiley-VCH, Weinheim (2012). Google Scholar
N. Krause. “Organogold chemistry”, in Organometallics in Synthesis – Fourth Manual B. H. Lipshutz (Ed.), pp. 429–540, Wiley, New York (2013).Google Scholar
F. D. Toste, V. Michelet, eds. Gold Catalysis: An Homogeneous Approach, Imperial College Press, London (2014). Google Scholar
Y. Zhang, J. Xue, Z. Xin, Z. Xie, Y. Li. Synlett 940 (2008). Google Scholar
A. Zhdanko, M. E. Maier. Eur. J. Org. Chem. 3411 (2014).Google Scholar
N. Krause, Ö. Aksin-Artok, M. Asikainen, V. Breker, C. Deutsch, J. Erdsack, H.-T. Fan, B. Gockel, S. Minkler, M. Poonoth, Y. Sawama, Y. Sawama, T. Sun, F. Volz, C. Winter. J. Organomet. Chem.704, 1 (2012). CrossrefGoogle Scholar
V. Belting. PhD Thesis, Dortmund University of Technology, 2011. Google Scholar
B. Wagner. PhD Thesis, Dortmund University of Technology, 2016. Google Scholar
K. Belger, N. Krause. Eur. J. Org. Chem. 220 (2015). Google Scholar
About the article
Published Online: 2016-06-01
Published in Print: 2016-04-01
Citation Information: Pure and Applied Chemistry, Volume 88, Issue 4, Pages 391–399, ISSN (Online) 1365-3075, ISSN (Print) 0033-4545, DOI: https://doi.org/10.1515/pac-2016-0406.http://creativecommons.org/licenses/by-nc-nd/4.0/.