SEARCH

SEARCH BY CITATION

Keywords:

  • alcohols;
  • alkynes;
  • isotopic labeling;
  • gold;
  • reaction mechanisms

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Acknowledgements
  7. Supporting Information

Under the conditions of dual activation catalysis with oxygen nucleophiles, β-substituted naphthalenes were obtained from 1,2-diethinyl arenes. Mechanistic studies, which include isotope labeling experiments, support that dual activation leads to β-substituted naphthalenes, whereas α-naphthalenes are formed by π activation only, and no gold acetylide or dual activation is involved in the formation of the α-substituted products. Additional experiments on substrates that led to dibenzopentalenes support these mechanistic insights.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Acknowledgements
  7. Supporting Information

Recently Zhang’s group1 and our group26 reported on a new reactivity pattern for gold-catalyzed reactions in which two gold fragments are used for substrate activation. One cationic gold fragment activates one π-bond of a diyne system, and the other gold molecule activates a second terminal alkyne by σ-coordination. This cooperation induces a cyclization that, dependent on the connecting tether of the diyne system, delivers carbene/vinylidene intermediates. These highly reactive intermediates have already been used for a series of fruitful transformations.16

In our first contribution on this topic, our group reported on a gold-catalyzed hydroarylating aromatization reaction.2 Depending on the reaction conditions, either α- or β-naphthalene derivatives were formed from simple diyne 1 by the incorporation of a benzene molecule, which was used as the reaction solvent (Scheme 1).

thumbnail image

Scheme 1. Selectivity control of the hydroarylating cyclization.

Download figure to PowerPoint

Based on these findings, we were curious to see if we could adapt this reaction to other nucleophiles.7 We envisioned that if stronger nucleophiles were used, it might be possible to perform the reaction in an inert solvent with only a slight excess of the nucleophile.

Results and Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Acknowledgements
  7. Supporting Information

To induce high β-selectivity, we applied σ,π-dinuclear propyne gold acetylides (dual activation catalysts; DACs) for our test reactions as these catalysts are ideal for dual activation catalysis cycles.8 As the test system, diyne 1 was used in combination with varying equivalents of methanol as the nucleophile (Table 1). Indeed it was possible to obtain a high β-selectivity for all of the applied ratios of nucleophile to substrate. If the amount of the protic nucleophile was higher than three equivalents, the competing formation of minor amounts of the α-substituted product was observed (entries 3–5). Next, we varied the counterions of the DACs (Table 2). Only a minor effect of the counterion was visible, but the tetrafluoroborate counterion was the best choice, and a perfect β-selectivity was observed for this counterion.

Table 1. Screening of the amount of nucleophile.inline image
EntryAmount of MeOHProduct yield [%][a]
 [equiv.]α-Substitutedβ-Substituted (4 a)
  1. [a] Yields determined after 3 d by GC–MS with hexamethylbenzene as internal standard.

11020
23traces44
35339
410340
550729
Table 2. Catalyst screening.inline image
EntryCounterion XProduct yield [%][a]
  α-Substitutedβ-Substituted (4 a)
  1. [a] Yields determined after 18 h by GC–MS with hexamethylbenzene as internal standard.

1SbF6443
2BF4047
3OTs029
4NNf2450
5PF6338
6NTf2243

We chose 1,2-dichloroethane (DCE) as the solvent for this reaction because many other solvents could be ruled out beforehand. Benzene, acetone, 1,4-dioxane, THF, and even nonpolar solvents such as cyclohexane and n-hexane reacted with the substrate under the reaction conditions and were, therefore, not suitable for this reaction. With the optimized conditions in hand (three equivalents of the nucleophile in combination with the tetrafluoroborate counterion), we turned our focus to the evaluation of the substrate scope (Table 3).

Table 3. Scope of the reaction.inline image
EntryProductYield of β-substituted product 4 [%][a]t
  1. [a] Isolated yields. [b] The diester was used instead of the dimethyl substrate.

1
Thumbnail image of
557 h
2
Thumbnail image of
243 d
3
Thumbnail image of
441 d
4
Thumbnail image of
4324 h
5
Thumbnail image of
4156 h
6[b]
Thumbnail image of
234 d

The isolated yield of β-methoxy naphthalene 4 a was only moderate (entry 1). An even lower yield was obtained by using phenol as a weaker nucleophile, which was accompanied by an increased reaction time. Thus we reverted back to other aliphatic alcohols that delivered yields in the range of 40 % regardless if branched or linear alcohols were applied (entries 3–5). If we changed the aromatic backbone, lower yields for the substrate dimethyl 4,5-diethynylphthalate (5), which bears electron-withdrawing ester moieties, were obtained (entry 6). Unfortunately, the substrate scope was quite limited, and amine derivatives (aniline, N-methylaniline, morpholine, and 1-(p-tolylsulfonyl)pyrrole) and carbon-based nucleophiles (dimethyl malonate) were not suitable for this transformation (for details see the Supporting Information). To prove the selectivity obtained, crystals of 4 c were grown that were suitable for XRD analysis.9 The molecular structure of the formed β-naphthol derivative 4 c is depicted in Figure 1.

thumbnail image

Figure 1. Solid-state structure of 4 c.

Download figure to PowerPoint

Based on our previous mechanistic investigation,2 we propose the mechanism depicted in Scheme 2. The reaction is initiated by the transfer of the two gold fragments from the DAC. The so-formed σ,π-activated starting material I can then undergo either a 5-endo-dig or a 6-endo-dig cyclization. Both of the pathways are reasonable and depend on the starting material. Both pathways have already been observed in the dual gold catalysis of diynes.6,10 In the case of the 5-endo-dig pathway, vinylidene II would be formed that, after the attack of methanol, would undergo a ring expansion to form the aurated naphthyl methyl ether IV, which is in equilibrium with gem-diaurated species V. The same species could also be formed through the 6-endo-dig pathway. In this case, after the initial cyclization, a fast 1,2-shift of the gold delivers a stabilized carbene/cation VII/VIII.11 The attack of methanol onto this species delivers the naphthyl methyl ether intermediate IV. The last step of the reaction cascade is catalyst transfer, which is crucial as otherwise the β-selectivity would be lost (no acetylide would be formed for the next catalytic cycle).

thumbnail image

Scheme 2. Mechanistic proposal.

Download figure to PowerPoint

To support our mechanistic proposal we conducted a deuterium labeling experiment by using three equivalents of [D4]MeOH as a the nucleophile (Scheme 3). Deuterium incorporation in the naphthalene backbone was exclusively found in the 1- (73 % D) and 4-positions (36 % D). This result strongly suggests that each or both pathways discussed in Scheme 2 can be involved. Deuterium incorporation in the 1-postion can be explained by the nucleophilic attack/protodeauration step from carbene/cation VII/VIII to the aurated naphthyl methyl ether IV if the reaction proceeds by the 6-endo-dig pathway. If the vinylidene pathway is favored, the deuterium is incorporated by the [D4]MeOH-induced ring-expansion step from vinylidene II to the aurated naphthyl methyl ether IV. The deuterium incorporation in the 4-position results from a [D4]MeOH-mediated protodeauration of the aurated naphthyl methyl ether IV to product 4 a. The lower deuterium incorporation in the 4-position with respect to the 1-position can be interpreted as a result of the competing catalyst transfer step that causes a protodeauration by the H atom of the diyne starting material.

thumbnail image

Scheme 3. Deuterium labeling experiment.

Download figure to PowerPoint

During the preparation of this manuscript, the group of Ohno reported on a cascade reaction of diynes that bear one phenyl-substituted alkyne in combination with a terminal alkyne. In their case, the exclusive formation of the α-substituted product was observed by using an external nucleophile (Scheme 4, top right).12 This is completely in line with our findings for the related hydroarylating aromatization, which was α-selective if no additive (a base or an organogold species) was present.2 However, there was a contradiction between the mechanistic proposal that was discussed for the formation of the α-substituted product and other preceding mechanistic insights. The authors speculate that for the formation of the α-substituted product an acetylide complex can be the key intermediate. This assumption is based on an experiment with gold acetylide 11 that was reported to deliver quantitative amounts of the α-substituted product 9 a if catalytic amounts of a cationic gold source were present (Scheme 4, bottom right).13 In addition, if Ohno et al. used a deuterated alkyne and 10 equivalents of EtOH as the nucleophile, almost a complete loss of deuterium was observed after the transformation, which was explained by an exchange process in which the intermediate gold acetylide was reprotonated. The involvement of an acetylide in the formation of the α-isomers is contradictory to our observation as exclusively β-substituted products are formed if additives for acetylide formation are present or if stoichiometric reactions that start from the preformed acetylides are performed. Furthermore, with exactly the same starting material in the absence of a nucleophile, our group was able to obtain benzopentalenes 7 (Scheme 4, top left). In addition, we could show that in this reaction gem-diaurated benzopentalenes were formed if an acetylide was used as the starting material in a stoichiometric experiment. For us it was not understandable that the same acetylide could give the α-substituted product in high yield. It seems clear that upon dual activation, intramolecular trapping should be favored, and, furthermore, in the case of an acetylide as the starting material, we would expect a β-substituted product to be formed if an external nucleophile could trap the intermediate.

thumbnail image

Scheme 4. Results of Ohno et al. and Hashmi et al. for the catalytic conversion of 8 and 11.

Download figure to PowerPoint

With this in mind, we performed the reaction with acetylide 11 and N-methylaniline under exactly the same conditions as those used by the Ohno group. In our experiment, we were able to detect 65 % of monoaurated benzopentalene 12 accompanied by only traces of the α-substituted product 9 a (<7 %; even <3 % if the reaction was performed in CH2Cl2 that was saturated with water; Scheme 4). The fact that only traces of the α-substituted product can be seen in the 1H NMR spectrum of the crude material is clear evidence that gold acetylides are not involved in the formation of the α-substituted product. Instead, if a gold acetylide is formed, direct trapping of the intermediate vinylidene occurs and the outcome is the expected benzopentalene product.

Indeed, only the α-substituted product 9 a can be obtained if the reaction is performed in the absence of an acetylide and by using N-methylaniline as the nucleophile. When we reproduced the reaction, we were able to obtain crystals suitable for XRD analysis, and the results deliver the final proof for the α-substitution (Figure 2).9 We assume that in this case, the reaction pathway is related to the α-pathway of the hydroarylating aromatization and the products are not derived from acetylides as starting materials. Both the cyclization by the free acetylene followed by a nucleophilic attack onto the so-formed aryl cation or the formation of an enamine/enol ether intermediate and a subsequent attack of the nucleophilic double bond onto the second alkyne are reasonable. Nevertheless, still no precise mechanistic picture exists for the formation of the α-substituted products.

thumbnail image

Figure 2. Solid-state structure of 9 a.

Download figure to PowerPoint

As Ohno et al. used ethanol as the nucleophile for both their screening reaction and most of their mechanistic investigations, we performed test reactions with this nucleophile too.

All their reported transformations were performed under an Ar atmosphere, therefore, we assumed that dry conditions (inert gas atmosphere, dry solvents, molecular sieves) were crucial. At first we encountered problems in trying to reproduce this reaction, but by variation of several parameters we found that trace amounts of water in the reaction mixture are essential for this reaction mode. The next reactions were, therefore, performed under air and without molecular sieves. First we reacted acetylide 11 with 2 mol % of IPrAuCl/AgOTf in the presence of 10 equivalents of ethanol. This gave monoaurated benzopentalene species 12, the product also obtained with N-methylaniline as the nucleophile (Scheme 4). The isolation of 12 in 58 % yield was possible despite the fact that the product was fairly unstable during column chromatography (Scheme 5).

thumbnail image

Scheme 5. Gold(I)-catalyzed conversion of 8 and 11.

Download figure to PowerPoint

With the purified product in hand, single crystals suitable for XRD analysis were obtained, which gave unambiguous proof of the formation of the monoaurated benzopentalene system.9 Interestingly, two different modifications were obtained in the solid state.9 Both solid-state molecular structures are depicted in Figure 3. In addition to the expected conformer with an orthogonal alignment of the two ligands, a conformer with all of the residues in one plane can be found in the other modification. The superimposed structures document this phenomenon well.

thumbnail image

Figure 3. Solid-state structure of 12.

Download figure to PowerPoint

Next, both groups together re-investigated the assumption of the Ohno group that the D/H exchange of the starting material (which can be monitored during the course of the reaction) and the accompanying loss of deuterium labeling in the product allows us to conclude that a gold acetylide is involved in the catalytic cycle. To evaluate if the acetylide can be protodemetalated, which would be crucial for H/D exchange, we heated acetylide 11 for 31 h at 50 °C but no traces of protodeauration were visible (Scheme 6). To exclude the involvement of a terminal alkyne in the exchange process, the same reaction was conducted in the presence of equimolar amounts of phenylacetylene but again no exchange was observed.

thumbnail image

Scheme 6. Reaction of 11 with ethanol.

Download figure to PowerPoint

As one might speculate that a second Au atom or a strong acid (which is indeed released during acetylide formation if no base is present) might be necessary for an exchange process, we subjected 11 to stoichiometric amounts of acid (both in normal CD2Cl2 and in CD2Cl2 that was saturated with water; Scheme 7).

thumbnail image

Scheme 7. Reaction of 11 with HBF4 etherate.

Download figure to PowerPoint

Indeed in all three cases the formation of the digold species 13 was observed in yields up to 84 % (referring to gold; yield determined by 1H NMR spectroscopy), whereas free diyne 8 was not seen in the 1H NMR spectra. This indicated that the protodemetalation of 11 is possible and then in a subsequent reaction 13 is formed by the dual activation pathway. If the same reaction was conducted in the presence of N-methylaniline as a potential nucleophile, the formation of significant amounts of the α-substituted product 9 a was observed (Scheme 7 b). This can be explained by the generation of free alkyne (in combination with free active catalyst) in the presence of a strong nucleophile that can deliver the α-substituted product through the non-acetylide pathway. In the presence of EtOH as a weaker nucleophile, hardly any α-substituted 9 b was obtained (Scheme 7 c), although 30 % of 13 was formed. Furthermore, we subjected σ,π-gem-diaurated phenylacetylene 14 (11 can cyclize through activation of the second alkyne) to stoichiometric amounts of acid at 50 °C (Scheme 8). It is not unexpected that no protodeauration takes place in this case as there are reports by Brown and Widenhofer who used exactly the reverse reaction for the preparation of σ,π-gold complexes.14

thumbnail image

Scheme 8. σ,π-gem-Diaurated phenylacetylene 14 does not react with acid.

Download figure to PowerPoint

No exchange could be monitored even at elevated temperatures. This shows that in contrast to 11 no protodemetalation is possible with σ,π-gem-diaurated complexes. The reason for the H/D exchange process reported by the Ohno group remains unclear. Two options are reasonable. One could be the direct H/D exchange via a gold π-coordinated alkyne, which increases the acidity of the proton at the alkyne. On the other hand, it is possible that an equilibrium between the free alkyne and the gold acetylide exists. This process is induced by the acid that is generated if no additive is present. This indicates that under the reaction conditions traces of acetylides might be formed, but for the formation of the α-substituted product the reverse reaction of the equilibrium must take place, which releases the free alkyne that then undergoes cyclization. In the presence of a dual catalyst or if one starts with an acetylide and catalytic amounts of a gold catalyst no acid is released and, therefore, the α-pathway is suppressed (Scheme 9).

thumbnail image

Scheme 9. Two possible pathways.

Download figure to PowerPoint

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Acknowledgements
  7. Supporting Information

We have shown that β-substituted naphthalene systems can be generated by a dual catalyzed gold reaction. The scope of the reaction is limited and only oxygen nucleophiles were suitable. Although for this kind of transformation acetylides are assumed to be key intermediates, we have shown that in the gold(I)-catalyzed formation of α-substituted naphthalene derivatives, which was previously reported by the Ohno group, gold(I) acetylenes 11 as true intermediate structures are unlikely. If 11 is formed, it follows the reaction mechanism to form benzopentalenes as reported by the Hashmi group in previous contributions. For the formation of the α-substituted product 9, free alkyne 8 is necessary in combination with free catalyst. Even if acetylides are present in the reaction, the equilibrium between free alkyne/cationic gold and acetylide/acid enables an α-pathway if a strong nucleophile is present.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Acknowledgements
  7. Supporting Information

The authors thank Umicore AG&Co. KG for the generous donation of gold salts.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Acknowledgements
  7. Supporting Information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
cctc_201300820_sm_miscellaneous_information.pdf5213Kmiscellaneous_information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.