Open Access Article
Jinhee
Kim
ab,
Mikyung
Sim
ab,
Namhoon
Kim
ab and
Sungwoo
Hong
*ba
aDepartment of Chemistry, Korea Advanced Institute of Science and Technology, Daejeon, 305-701, Korea
bCenter for Catalytic Hydrocarbon Functionalization, Institute for Basic Science (IBS), Daejeon, 305-701, Korea. E-mail: hongorg@kaist.ac.kr
First published on 16th April 2015
The systematic investigation of substrate-bound α-amino acid auxiliaries has resulted in catalytic asymmetric C–H functionalization of cyclopropanes enabled by amino acid amides as chiral bidentate directing groups. The use of an Ile-NH2 auxiliary embedded in the substrate provided excellent levels of asymmetric induction (diastereomeric ratio of up to 72
:
1) in the Pd(II)-catalyzed β-methylene C(sp3)–H bond activation of cyclopropanes and cross-coupling with aryl iodides.
Elegant examples of the diastereoselective functionalization of unactivated C(sp3)–H in chiral substrates were revealed by the Corey,3b Daugulis3q and Chen groups.3d,3m Moreover, remarkable achievements in catalytic asymmetric C–H activation have been demonstrated using a chiral auxiliary8 and a chiral ligand.9 We hypothesized that an appropriate chiral bidentate directing group embedded in the substrate could induce high levels of stereocontrol during C–H functionalization via a steric repulsion model. As an outgrowth of these studies, we selected α-amino acid moieties as chiral bidentate directing groups in the context of the diastereotopic β-methylene C(sp3)–H bond functionalization/arylation of cyclopropanes, enabling the stereoselective installation of an aryl group and the construction of new stereogenic centers (Scheme 1). Cyclopropanes are a privileged class of structures found in many biologically active molecules.10 In this regard, the transition-metal-catalyzed asymmetric C–H arylation of cyclopropanes has been the focus of great research interest.9,11 Moreover, functionalized cyclopropanes bearing amino acid moieties are widely used as a conformationally restricting linker in medicinal chemistry.10d,e Herein, we report the first example of chiral bidentate auxiliary-controlled diastereoselective C(sp3)–H bond activation/cross-coupling of cyclopropanes. From a conceptual viewpoint, the ability of a substrate-bound α-amino acid auxiliary to promote diastereoselective C–H functionalization is intriguing because the amino acid moiety derived from readily available chiral pools not only plays the role of a bidentate directing group, but also of a chiral auxiliary to provide efficient stereocontrol during C(sp3)–H bond functionalization. As a consequence, high levels of asymmetric induction could be expected in the C(sp3)–H functionalization of cyclopropanes controlled by a chiral bidentate directing group.
:
1). This preliminary work prompted us to scrutinize additional types of amino acid amides as stereogenic directing groups. Further investigations revealed that having the Val–NH2 group (1f) was crucial for obtaining a higher reactivity and diastereoselectivity (entry 6, 60%, d.r = 8.5
:
1). The Val–NMe2 group (1e) attached to the substrate led to a complete loss of reactivity (entry 5), indicating the favorable property of an NH2 motif on coordination in the active complex. We therefore decided to employ Val–NH2 (1f) as a directing group for further optimization; representative catalytic systems for the cross-coupling are listed in Table 1 (entries 6–11).12 In general, an approximate 4.3
:
1 mixture of mono- and diarylated products were obtained with substrate 1f, containing Val–NH2, at 100 °C. Attempts to improve the conversion at a higher temperature (120 °C, entry 7) resulted in an increased proportion of the diarylated product (81%, mono/di = 2.5
:
1, d.r = 6.3
:
1) with mixtures of polyarylated by-products13 (∼10%). Several solvents were also evaluated and the use of t-amyl-OH as the solvent was necessary to achieve a higher conversion. The use of AgOAc (entry 9), which was employed previously in arylation reactions,10d,e provided a lower product yield (20%) along with the recovery of the remaining starting material (58%). Alternative arylating coupling partners, such as diphenyliodonium salt3o failed to improve the reaction efficiency (entry 10). With the use of 1.5 equiv. of iodobenzene, the reaction reached full conversion to provide the mono- and diarylated products in yields of 71 and 16%, respectively, with only a negligible amount of the undesired by-products. Thus, the best results were obtained with the substrate (1 equiv.) with iodobenzene (1.5 equiv.), Pd(OAc)2 (10 mol%), and K2CO3 (2 equiv.) in t-amyl-OH (0.5 M) at 100 °C to give the monoarylation product (71%) with high diastereoselectivity (d.r = 8.6
:
1) (entry 11). The structure of the major diastereomer 2a (Val–NH2 derivative) was unambiguously verified by H NMR and X-ray crystallographic analyses (Fig. 1).
| Entry | R | PhX (equiv.) | Base (equiv.) | Temp. (°C) | Yieldb (d.r)c |
|---|---|---|---|---|---|
a Substrate (1.0 equiv.), Pd(OAc)2 (10 mol%), K2CO3 (2.5 equiv.), and ArX in t-amyl-OH (0.5 M).
b Isolated yields of monoarylation products.
c The d.r was determined by HPLC analysis.
d Mono- and diarylation products (4.5 : 1) were observed by NMR analysis.
|
|||||
| 1 |
|
PhI (3) | K2CO3 (2.5) | 100 | Trace |
| 2 |
|
PhI (3) | K2CO3 (2.5) | 100 | NR |
| 3 |
|
PhI (3) | K2CO3 (2.5) | 100 | NR |
| 4 |
|
PhI (3) | K2CO3 (2.5) | 100 | 26% (8.1 : 1) |
| 5 |
|
PhI (3) | K2CO3 (2.5) | 100 | NR |
| 6 |
|
PhI (3) | K2CO3 (2.5) | 100 | 60% (8.5 : 1) |
| 7 | PhI (3) | K2CO3 (2.5) | 120 | 58% (6.3 : 1) |
|
| 8 | PhI (3) | K2CO3 (2.5) | 80 | 12% | |
| 9 | PhI (4) | AgOAc (2.5) | 100 | 20% | |
| 10 | Ph2IOTf (2) | K2CO3 (3) | 100 | 33% | |
| 11d | PhI (1.5) | K2CO3 (2) | 100 | 71% (8.6 : 1) |
|
Having identified that the N-nonsubstituted –CONH2 moiety was crucial for achieving both high reactivity and diastereoselectivity, we further explored the effect of steric bulk at the α-position of amino acid amides on the selectivity. For this purpose, substrates bearing a range of R substituents were examined and chiral auxiliaries featuring sterically bulky side chains appear to be essential for achieving high levels of stereoselectivity (Scheme 2). Importantly, the replacement of an isopropyl group (d.r = 8.6
:
1) at the α-position of the amino acid amide with sterically demanding isobutyl and t-butyl groups produced the corresponding products 3a and 3b with higher diastereoselectivity (d.r = 10.2
:
1 for 3a, d.r = 13.7
:
1 for 3b). Further studies revealed that the isobutyl and t-butyl groups present on the substrates gave superior results to those with benzyloxy ethyl (3c), phenyl (3d), cyclohexyl (3e), or benzyl (3f) groups. Of the various chiral auxiliaries that were tested, the Ile-NH2 moiety (3a) was selected as a bidentate directing group after considering both yield and diastereoselectivity. In order to gain insight into the influence of the diarylation process on the diastereoselectivity, a reaction profile was performed by monitoring the conversion to 3a under standard conditions (see Table S4 in the ESI† for more details). An increase of d.r from 7.2
:
1 (after 1 h) to 10.2
:
1 (after 24 h) was observed, which suggests that the favorable second arylation of the minor stereoisomer may contribute, in part, to the overall diastereoselectivity observed.
![]() | ||
Scheme 2 Influence of the side chain on C–H arylation. Substrate (1.0 equiv.), Pd(OAc)2 (10 mol%), K2CO3 (2.0 equiv.), and PhI (1.5 equiv.) in t-amyl-OH (0.5 M) at 100 °C for 24 h: isolated yields of monoarylation products after purification by silica gel chromatography. The d.r was determined by HPLC analysis. In general, mixtures of mono- and diarylation products (4.0–4.5 : 1) were observed by NMR analysis. aAfter 1 h: NMR yield of monoarylation products (see Table S4 in the ESI† for more details). | ||
The directing group is removable and the hydrolysis of the amino acid amide group of 3a occurred smoothly under acidic conditions to afford arylated cyclopropanecarboxylic acid (81%) with the conservation of the stereogenic centers (eqn (1)).14,15
![]() | (1) |
Having established that the Pd(II)-catalyzed highly diastereoselective C–H arylation of cyclopropanes was enabled by the Ile-NH2 directing group, the scope of aryl iodides was studied with a range of functional groups such as phenyl, methoxy, ketone, ester, alkyl, halides, and trifluoromethyl, which could be used as a useful synthetic handle for the further transformations. We were delighted to observe that substitution with both electron-donating (Me–, tBu–, and OMe–) and electron-withdrawing groups (F–, Cl–, Br–, CF3–, CO2Me–, and COMe–) on the aryl iodides were viable under the reaction conditions to afford the corresponding products with high levels of diastereoselectivity, summarized in Scheme 3. In addition, the major isomer could be easily isolated by silica gel chromatography. Outstanding diastereoselectivities could be achieved in the reactions of 3,5-disubstituted iodobenzenes (d.r = 30.1
:
1 for 3k, d.r = 71.5
:
1 for 3l). The chemoselective coupling with bromo iodobenzene afforded the synthetically important 3n, of use for further synthetic elaboration. The position of the substituents in the iodobenzene did not show much change to the reactivity. Notably, aryl iodides with ortho-substituents, such as OMe, Cl, and F were also compatible with the reaction conditions to afford the corresponding products (3w–y) with good yields (67–73%). Expanding the scope to nitro-substituted iodobenzenes was also possible, but the products were obtained in only around 30% yields (3u and 3v).
Although severe peak overlap occurs in the NMR spectra of two diastereotopic hydrogens, a significant level of the di-deuterated product (>40%) was observed when the reaction mixture was treated with D2O under the optimized conditions and in the absence of aryl iodide (Scheme 4).12 This controlled H/D exchange study implies a facile C–H activation process for both diastereotopic hydrogens of the cyclopropanes.
We proposed one of the possible reaction mechanisms involving Ile-NH2 auxiliary-controlled asymmetric C–H functionalization in which a Pd(II)/Pd(IV)16 catalytic cycle is invoked (Scheme 5). The coordination of two nitrogen atoms to the Pd catalyst generates a palladium amide I. Subsequently, the cleavage of the C(sp3)–H bond on the cyclopropane through a concerted metalation deprotonation (CMD) process produces the palladacycle complexes II and III, which may exist in equilibrium.17 The diastereoselectivity of the arylation could occur either at the C–H activation step or at the oxidative addition step. At this stage, the diastereoselectivity can be assumed to be established at the oxidative addition step under Curtin–Hammett control18 because the C–H activation process of both diastereotopic hydrogens is relatively facile (Scheme 4). The oxidative addition of the aryl iodide to II affords Pd(IV) intermediate IV,19 which then undergoes reductive elimination followed by protonation leading to the formation of the arylated product. On the other hand, the oxidative addition process for the insertion of an aryl iodide into Pd complex III appears to be less feasible because the R substituent and the methylene group of cyclopropane cause blocking of the top and bottom faces.
![]() | ||
| Scheme 5 Mechanistic considerations for Pd-catalyzed diastereoselective C–H activation of cyclopropanes. | ||
Furthermore, we preliminarily investigated the diastereoselective C–H alkynylation and observed that 1f reacted with bromoalkyne under the same reaction conditions,20 resulting in the single diastereomer 4 in a 53% yield along with 44% of unreacted starting material (Scheme 6).
Footnote |
| † Electronic supplementary information (ESI) available. CCDC 1053232. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c5sc01137j |
| This journal is © The Royal Society of Chemistry 2015 |