Dedicated to Prof. Barry M. Trost
Key words
1,2-dihydroquinoline - 1,2,3,4-tetrahydroquinoline - copper hydride - asymmetric catalysis - hydrosilylation - dearomatization
Silicon is regarded as a bio-isostere of carbon in clinical studies due to their similarity.[1] In fact, the introduction of silicon-containing groups to known bioactive compounds provides opportunities to alter their properties (Scheme [1]).[2] For example, the larger covalent radius of silicon and the longer Si–C covalent bond lengths (compared to C–C bonds) may lead to different conformations of the biomolecules, thus providing beneficial influence on their biological activities.[3] Furthermore, introducing the silyl group to biomolecules may adjust their pharmacokinetic pathway.[4]
[5] In this regard, numerous methods were developed for the synthesis of potential bioactive silicon-containing compounds, which could be roughly classified into two types. One is the installation of a silicon-containing moiety on known lead compounds or drugs[6] (e.g., Camptothecin versus Karenitecin)[6b] albeit with slightly low efficiency.[1d]
[7] The second is the introduction of silyl groups or carbon-to-silicon switch on a bioactive skeleton.[8] However, the latter remains underexplored and most of the reported studies are restricted to the modification of amino acids.[9] Despite the considerable progress, incorporation of silicon moieties on diverse bioactive scaffolds is still in great demand.
Scheme 1 Application of silicon in pharmaceutical chemistry
1,2,3,4-Tetrahydroquinolines (THQs) represent an important bioactive skeleton because of their ubiquitous presence in pharmaceuticals and natural products.[10] Diverse methods have been developed for the synthesis of chiral THQs such as assorted cyclizations (e.g., Povarov reaction,[11] Michael addition,[12] C–H bond functionalization[13]),[14] kinetic resolution[15] and dearomatization. In particular, asymmetric dearomatization of quinolines constitutes a straightforward approach. By utilizing (transfer) hydrogenation[16] or the Reissert type reaction,[17] various enantioenriched THQs were obtained from readily available quinolines. Recently, a stepwise reduction of quinolines and asymmetric catalytic transformation of the generated dihydroquinolines emerged as an attractive method to access chiral THQs with no need for preactivation of quinoline substrates.[18] Thus the combination of this strategy with the introduction of silyl groups would be an appealing approach for the synthesis of silyl substituted THQs.[19] Compared with other methods of forging C(sp3)–Si bonds,[20] transition-metal-catalyzed asymmetric hydrosilylation of unsaturated compounds[21] represents a direct and atom-economic approach.[22] In this regard, we envisioned that copper-catalyzed asymmetric hydrosilylation of 1,2-dihydroquinolines would efficiently introduce a C–Si bond at the 4-position of THQs.[23] Herein, we report the results of this study.
At the outset, N-CO2Me 1,2-dihydroquinoline (1a) was chosen as the substrate in the hydrosilylation reaction (Table [1]). Considering the relatively high reactivity of arylsilanes over alkylsilanes, we chose diphenylsilane for the initial attempt. Compound 1a was treated with 3 equivalents of diphenylsilane in the presence of the catalyst derived from copper(II) acetate and L1 at 40 ° under neat conditions for 36 h. To our delight, the desired product 2a was afforded in 56% yield and 87% ee (entry 1). Notably, the addition of a mono-phosphine as the secondary ligand[24] resulted in an improved yield (entries 2–5; see the Supporting Information for more details), and (p-tolyl)3P gave the optimal results (86% yield and 87% ee; entry 4).[25] Subsequent screening of chiral ligands revealed that L1 is the most effective ligand (entries 6–11).
Table 1 Optimization of the Reaction Conditionsa
|
Entry
|
Secondary ligand
|
Ligand
|
Yield (%)b
|
ee (%)c
|
1
|
none
|
L1
|
56
|
87
|
2
|
PPh3
|
L1
|
84
|
86
|
3
|
(4-MeOC6H4)3P
|
L1
|
64
|
85
|
4
|
(p-tolyl)3P
|
L1
|
86
|
87
|
5
|
PCy3
|
L1
|
80
|
56
|
6
|
(p-tolyl)3P
|
L2
|
22
|
1
|
7
|
(p-tolyl)3P
|
L3
|
8
|
1
|
8
|
(p-tolyl)3P
|
L4
|
0
|
n.a.
|
9
|
(p-tolyl)3P
|
L5
|
0
|
n.a.
|
10
|
(p-tolyl)3P
|
L6
|
0
|
n.a.
|
11
|
(p-tolyl)3P
|
L7
|
0
|
n.a.
|
a Reaction conditions: Cu(OAc)2 (0.010 mmol), ligand (0.011 mmol), secondary ligand (0.022 mmol, if used), 1a (0.2 mmol) and diphenylsilane (0.6 mmol) were stirred at 40 °C under neat conditions for 36 h.
b Isolated yield.
c The ee value was determined based on HPLC analysis; n.a. = not applicable.
With the optimal conditions in hand, the substrate scope was then explored (Scheme [2]; see also the Supporting Information and Procedure A).[26] The substrates with various N-protecting groups (Ac, CO2Bn, CO2
i
Bu) gave the desired products with good yields and enantioselectivities (2b–d, 74–83% yields, 83–90% ee). A series of substituents at the 6-position of 1,2-dihydroquinolines were explored. A moderate yield and poor enantioselectivity were observed for the bromo-bearing substrate (2e, 50% yield, 46% ee). The substrate with an electron-withdrawing group (CO2Me) worked well, affording the desired THQ in 87% yield and 89% ee (2f). 6-Thienyl-1,2-dihydroquinoline was transformed into its corresponding product 2g in moderate yield with slightly decreased ee value (47% yield, 76% ee). The THQs with electron-donating groups (OMe and SMe) were obtained with good results (2h, 87% yield, 94% ee; 2i, 82% yield, 86% ee). Substituent effects were also investigated for the 7-methyl (2j) and 7-methoxy (2k) substrates, giving 77% yield with 82% ee and 62% yield with 82% ee, respectively. Subsequently, by utilizing phenylsilane as the silyl reagent, the desired 4-silyl THQs were generated and an extra Tamao oxidation was performed in a one-pot fashion, yielding 4-hydroxy THQs (Scheme [3]; see also the Supporting Information and Procedure B).[27] The N-CO2Me and N-CO2Bn 1,2-dihydroquinolines were well tolerated, leading to the desired products 3a
[28] and 3b in 63% yield with 89% ee and 73% yield with 88% ee, respectively. 1,2-Dihydroquinolines bearing varied substituents (6-OMe, 6-Ph and 7-Me) reacted smoothly with phenylsilane, and moderate yields with good enantioselectivities were obtained (3c–e, 50–68% yields, 85–89% ee). Notably, other silanes such as Et2SiH2 and Et2MeSiH were also tested, but failed to give any desired product.
To demonstrate the practicality of this protocol, we next performed scale-up reactions (Scheme [4]). The hydrosilylation of 1a with diphenylsilane at 5 mmol scale under the standard conditions gave an improved yield and slightly decreased enantioselectivity (eq. 1, 1.76 g, 94% yield, 85% ee). The one-pot hydrosilylation/Tamao oxidation reaction occurred smoothly with 2.5 mol% catalyst loading (eq. 2, 168 mg, 81% yield, 89% ee). The 4-silyl THQ product 2a could be oxidized to the desired silanol in 83% yield with 85% ee (Scheme [5]). The absolute configuration of 4 was determined to be R by the X-ray crystallographic analysis of its optically pure single crystal and the absolute configurations of the products 2 and 3 were assigned based on the assignment of 4.[29]
Scheme 2 Substrate scope for the asymmetric hydrosilylation. Reagents and conditions: Cu(OAc)2 (0.010 mmol), (R,R)-Ph-BPE (0.011 mmol), (p-tolyl)3P (0.022 mmol), 1 (0.2 mmol) and diphenylsilane (0.6 mmol) were stirred at 40 °C under neat condition for 36 h. Isolated yields are reported and the ee value was determined based on HPLC or SFC analysis. a 5 equiv of silane were used. b The reaction time was 48 h.
Scheme 3 Substrate scope for the asymmetric hydrosilylation and Tamao oxidation. Reagents and conditions: Cu(OAc)2 (0.010 mmol), (R,R)-Ph-BPE (0.011 mmol), (p-tolyl)3P (0.022 mmol), 1 (0.2 mmol) and phenylsilane (0.6 mmol) were stirred at 40 °C under neat conditions for 36 h. Then KF (0.8 mmol), KHCO3 (0.8 mmol), K2EDTA·(H2O)2 (0.2 mmol), MeOH (1.2 mL) and H2O2 (1.8 mmol) were added and stirred at r.t. in THF (1.2 mL) for 20 h. Isolated yields are given and the ee values were determined based on HPLC or SFC analysis.
A deuterium experiment utilizing Ph2SiD2 as the hydride source revealed that hydrogen atoms at both the 3-position and in the silyl group were deuterated, which demonstrated the excellent atom-economy of this protocol (Scheme [6]). A plausible mechanism was thus proposed as exemplified by the reaction between 1a and diphenylsilane (Scheme [7]). Ligated copper hydride (LCuH) is generated in situ from the copper(II) acetate, the ligand (L) and diphenylsilane. The activated species then inserts into 1a with formation of intermediate A containing the stereogenic center with a C–Cu bond. Subsequent stereoretentive σ-metathesis between A and another silane molecule results in the desired product 2a and the regeneration of LCuH.[30]
Scheme 4 Scale-up reactions
Scheme 5 Transformation into product 4
Scheme 6 Deuterium experiment
Scheme 7 Plausible mechanism
In conclusion, a copper(II) acetate/(R,R)-Ph-BPE/(p-tolyl)3P catalyzed asymmetric hydrosilylation of 1,2-dihydroquinolines with hydrosilanes was developed. Various 4-silyl and 4-hydroxy 1,2,3,4-tetrahydroquinolines were obtained with good enantioselectivities. The enantioselective incorporation of a silyl group on the tetrahydroquinoline skeleton might find application in medicinal chemistry.