Aakarsh
Saxena
and
Hon Wai
Lam
*
EaStCHEM, School of Chemistry, University of Edinburgh, Joseph Black Building, The King's Buildings, West Mains Road, Edinburgh, EH9 3JJ, United Kingdom. E-mail: h.lam@ed.ac.uk:; Tel: +44-131-650-4831
First published on 8th September 2011
β-Substituted alkenyl-para-nitroarenes, an unexplored substrate class for catalytic asymmetric addition reactions, undergo highly enantioselective rhodium-catalyzed arylations with arylboronic acids in the presence of a dibenzylamide-containing chiral diene ligand. One example of the asymmetric arylation of an alkenyl-p-cyano-m-(trifluoromethyl)benzene is also presented.
Our group has demonstrated that heteroarenes containing a suitably placed C
N moiety are able to activate alkenes2 towards enantioselective copper-catalyzed reductions3 and rhodium-catalyzed arylations,4 while Bernadi, Adamo, and co-workers have developed asymmetric additions of nitroalkanes to 4-nitro-5-styrylisoxazoles.5 For alkenes conjugated to simple arenes (which in general provide only minimal activation), highly reactive organometallic reagents are usually required.6 A number of groups have reported stoichiometric or catalytic enantioselective carbolithiations7,8 of various alkenylarenes9–11 mediated by (–)-sparteine9,10 or a (+)-sparteine surrogate.11 Although these carbolithiations work well, the development of processes that proceed under milder conditions, employing organometallic reagents that exhibit greater functional group compatibility, represents an unmet need. In this paper, the catalytic enantioselective addition of arylboronic acids to alkenes conjugated to electron-deficient arenes is described.
The rhodium-catalyzed asymmetric 1,4-addition of arylboron compounds to β-substituted electron-deficient alkenes is now a well-established method for the preparation of chiral compounds.12–14 Since the initial discovery of enones as substrates for these reactions,12 subsequent efforts have extended the scope of the acceptor15 to alkenes conjugated to a range of common electron-withdrawing groups.16–22 More recently, less common activating groups have been employed. In addition to our report on asymmetric arylations of alkenylheteroarenes,4 which builds upon work by Lautens and co-workers describing non-enantioselective additions of boronic acids to vinylazines,23 Sasaki and Hayashi have disclosed the asymmetric arylation of borylalkenes.24 However, rhodium-catalyzed addition of arylboron compounds to alkenylarenes has not, to our knowledge, been described. Instead, simple styrenes (where activation of the alkene is minimal) were shown by the Lautens group to undergo Heck-type reactions under aqueous conditions using water-soluble phosphine ligands (eqn (1)).23
![]() | (1) |
![]() | (2) |
It occurred to us that placement of the strongly electron-withdrawing nitro group at the para-position of the arene might lead to sufficient polarization of the alkene to the point where addition products 2, rather than Heck-type products, would form, even for 1,2-disubstituted alkenes (eqn (2)). Although nitroalkenes have been successfully employed in myriad additions of carbon nucleophiles,19,25 the analogous reactions of their phenylogous counterparts 1 are extremely rare,6c–h and no asymmetric reactions have been reported. Therefore, the successful realization of the reactions depicted in eqn (2) was an attractive goal and would set the stage for the use of this under-exploited class of electrophiles in other catalytic enantioselective addition reactions.
Our initial experiments focused upon alkenyl-p-nitroarene 1a as a test substrate (Table 1). As a preliminary gauge of reactivity, the addition of PhB(OH)2 to 1a was performed using [Rh(cod)Cl]2 (2.5 mol%) and KOH (2.5 equiv.) in dioxane/H2O at 80 °C under microwave (μw) irradiation26 for 30 min. This experiment resulted in 42% conversion into rac-2a (entry 1). Next, the use of chiral ligands was evaluated in combination with [Rh(C2H4)2Cl]2 as a precatalyst to assess whether 2a could be obtained with improved conversions and in high enantioselectivity. Chiral diene ligands have been shown to provide excellent results in asymmetric 1,4-arylation reactions27–29 and, in view of the success obtained with secondary amide-containing ligand L130 in our study of the asymmetric arylation of alkenylheteroarenes,4 this diene was evaluated first. Although L1 did lead to 2a in 97% ee, the conversion was only 35% (entry 2). Increasing the temperature to 120 °C did increase the conversion with only a slight impact upon enantioselection (95% ee), but appreciable starting material remained (entry 3). Additional amide-containing chiral dienes were then investigated. The enantioselectivity remained high with ligand L2 that lacks the pyrrole on the cyclohexyl ring, but the conversion was low (entry 4). Ligand L34 containing a morpholine amide provided improved conversion (76%) at 80 °C, but the product was formed in only 70% ee (entry 5). Ligands L4 and L5 containing tertiary amides gave improved results (entries 6 and 7), with dibenzylamide-containing ligand L5 giving the product in >95% conversion, 92% isolated yield, and 95% ee (entry 7). In contrast, ligand L6 containing only one benzyl group on the amide nitrogen atom afforded inferior results (entry 8), further suggesting that under these conditions, a tertiary amide in the ligand is beneficial for high conversion. Finally, (R)-BINAP (L7) was tested for comparison and although the enantioselectivity was high, the reaction did not go to completion (entry 9). On the basis of these results, ligand L5 was selected for further study.
a Reactions were conducted using 0.20 mmol of 1a in dioxane (0.5 mL) and H2O (0.1 mL).
b
Determined by 1H NMR analysis of the unpurified reaction mixtures. c Determined by HPLC analysis on a chiral stationary phase. d [Rh(cod)Cl]2 was used in place of [Rh(C2H4)2Cl]2, without an additional chiral ligand. e Reaction conducted at 120 °C for 30 min. f Product 2a was isolated in 92% yield. |
|---|
|
Next, the addition of a range of arylboronic acids to various alkenyl-p-nitroarenes was investigated (Table 2), and the enantioselectivity of the reaction was, in most cases, high (84–97% ee). In addition to a p-nitrophenyl group (entries 1–12), other arenes that provide effective activation in this process include o-fluoro-p-nitrophenyl (entry 14), m-methyl-p-nitrophenyl (entry 15), m-carbomethoxy-p-nitrophenyl (entry 16), and p-nitro-m-(trifluoromethyl)phenyl (entry 17). The reaction is not limited to alkenyl-p-nitrobenzenes; substrate 1k containing a 4-nitronaphthyl group also underwent arylation to provide 2r, though the yield and enantioselectivity were somewhat diminished with this sterically more demanding substrate (entry 18). The range of tolerated substituents at the β-position of the alkene include simple linear alkyl groups (entries 1–9 and 14–18), a cyclopropyl group (entry 10), an allyl ether (entry 11) and an allyl amine (entry 12). However, a β-aryl group was found to inhibit the reaction (entry 13). Regarding the scope of the nucleophile, arylboronic acids containing methyl, halogen, or methoxy substituents were competent reaction partners in this process. The reaction of sterically demanding 2-methylphenylboronic acid with substrate 1b provided 2f in 97% ee, though in a modest 61% yield (entry 6). Thermal heating is as effective as microwave heating, as evidenced by a reaction conducted under otherwise identical conditions (entry 1, values in parentheses). Furthermore, thermal heating was employed in the addition of phenylboronic acid to 1b on a 1.0 mmol scale with 1.25 mol% of [Rh(C2H4)2Cl]2 and 3 mol% of L5 at 80 °C for 1 h, which provided 2c in 83% yield and 95% ee (entry 3).
a Unless otherwise stated, reactions were conducted using 0.20 mmol of 1a–1k. Cited yields are of isolated material. Enantiomeric excesses were determined by chiral HPLC analysis.
b
Values in parentheses refer to a reaction conducted under thermal heating under otherwise identical conditions. c Reaction performed using 1.0 mmol of 1b at 80 °C under thermal heating for 1 h, using 2.5 mol% of Rh and 3 mol% of L5. d Reaction time was 1 h. e Reaction performed using 1.0 mmol of 1h. |
|---|
|
An additional demonstration of the reaction scope is provided in eqn (3), where substrate 3 containing a β-trimethylsilyl substituent underwent arylation in 57% yield and 91% ee.31
![]() | (3) |
To further test the utility of this process, a preparative-scale reaction was performed using substrate 4 (5.0 mmol) containing an oxygenated alkyl substituent at the β-position (Scheme 1). This experiment provided 2t in 88% yield and 93% ee. In addition, reduction of the nitro group of 2t, followed by tosylation of the resulting amine 5, provided sulfonamide 6 in 91% yield over two steps, the absolute stereochemistry of which was determined by single crystal X-ray analysis (Fig. 1).‡32 The sense of enantioinduction observed using ligand L5 is consistent with the stereochemical model proposed for previously reported examples of arylation of acyclic electron-deficient alkenes using structurally similar chiral dienes.4,30 In this model, the rhodium–aryl bond is situated trans to the more electron-deficient alkene, and binding of the alkenylnitroarene occurs in a manner that minimizes unfavorable steric interactions (Fig. 2).
![]() | ||
| Scheme 1 Larger-scale arylation of 4 and subsequent elaboration. | ||
![]() | ||
| Fig. 1 An ORTEP drawing of 6 with ellipsoids set at 50% probability. | ||
![]() | ||
| Fig. 2 A model for stereochemical induction. | ||
Nitroarenes are well-known to undergo a range of valuable reactions, making them versatile intermediates in the preparation of dyes, pharmaceuticals, and other functional compounds.33 To demonstrate the synthetic utility of the arylation products described herein, 2o was smoothly converted into indole 7 in 67% yield by treatment with vinylmagnesium bromide according to the method of Bartoli and co-workers (eqn (4)).34,35
![]() | (4) |
Further experiments provided insights into the structural features required in the substrate for the reaction to proceed under the present conditions. Substrates 8 and 9 containing m-nitrophenyl and o-nitrophenyl groups, respectively, did not provide the desired arylation products (Fig. 3).
![]() | ||
| Fig. 3 Unreactive substrates. | ||
While the lack of reactivity of 8 is not surprising given that the nitro group is not conjugated with the alkene,6d the failure of 9 to undergo arylation was somewhat unexpected, given that o-nitrostyrene has been shown to react smoothly with a variety of active methylene compounds under basic conditions.6d The attempted arylation of 9 using a stoichiometric quantity of the rhodium-ligand complex also provided no evidence of the desired product, suggesting that the problem is one of reactivity rather than catalyst turnover. The addition of 10 mol% of substrate 9 to a repeat of the reaction of Table 2, entry 1 under otherwise identical conditions led to the formation of 2a in >95% conversion and 94% ee, further suggesting that 9 does not poison the catalyst. Exactly how the o-nitro group in 9 inhibits the carborhodation step in the mechanism of rhodium-catalyzed addition of arylboronic acids to electron-deficient alkenes36 is not known at this time.
Nevertheless, the powerful effect of a p-nitro group allowed us to address a problem discovered during our recent study of enantioselective rhodium-catalyzed additions of arylboronic acids to alkenylheteroarenes, which identified a 2-pyridyl group as providing insufficient activation of an adjacent alkene for arylation to proceed efficiently.4 Gratifyingly, 2-alkenylpyridine 10 containing a 5-nitro group underwent arylation in high yield and enantioselectivity (eqn (5)).
![]() | (5) |
Finally, efforts to employ alkenylbenzene substrates containing a single para-electron-withdrawing substituent other than a nitro group, such as acetyl, nitrile, or methanesulfonyl, were unsuccessful with only low conversions into mixtures of identified products being observed. However, substrate 12, containing a p-cyano-m-(trifluoromethyl)phenyl group, did undergo arylation in 59% yield and 84% ee in the presence of 10 mol% of the rhodium–chiral diene complex after 1.5 h (eqn (6)).
![]() | (6) |
In contrast, no reaction was observed using (R)-BINAP (L7) as the ligand. The result of eqn (6) suggests that there is scope to increase the range of electron-deficient arenes that can be used as activating groups and future developments in this area may rest upon the identification of more active catalysts and/or improved reaction conditions.
Footnotes |
| † Dedicated to the memory of Hamish McNab. |
| ‡ Electronic supplementary information (ESI) available: Experimental procedures, spectroscopic data for new compounds, and crystallographic data. CCDC reference number 809081. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c1sc00521a |
| This journal is © The Royal Society of Chemistry 2011 |