Trityl Cation-Catalyzed Hosomi-Sakurai Reaction of Allylsilane with β,γ-Unsaturated α-Ketoester to Form γ,γ-Disubstituted α-Ketoesters.

(Ph3C)[BPh(F)4]-catalyzed Hosomi-Sakurai allylation of allylsilanes with β,γ-unsaturated α-ketoesters has been developed to give γ,γ-disubstituted α-ketoesters in high yields with excellent chemoselectivity. Preliminary mechanistic studies suggest that trityl cation dominates the catalysis, while the silyl cation plays a minor role.

1. Introduction

α-Ketoesters are important synthons [1,2,3] and can be transformed into a variety of building blocks, which have found wide utility in natural products synthesis. As shown in Scheme 1, a Bi(OTf)3-catalzyzed intermolecular cascade annulation of α-ketoesters with alkynols has been developed to construct γ-spiroketal-γ-lactones [4], a core structure in massarinoline A [5]. α-Ketoesters can also react with α-ketoacids to form isotetronic acids [6], a core structure in aspernolide A [7], by an asymmetric aldol/lactonization/enolization reaction. In this regard, development of new methods enabling an efficient synthesis of structurally diverse α-ketoesters is highly desirable.

Scheme 1

Synthetic utility of α-ketoesters in the synthesis of natural products. “cat*” refers to chiral catalysts.

Trityl tetrakis (pentafluorophenyl) borate [(Ph3C)[BPh(F)4]] [8] is well-known for providing stable and easily available carbocations. Since the pioneering work of Mukaiyama and co-workers [9], the use of trityl cations as Lewis acid catalysts has been explored in Mukaiyama aldol reactions [10,11,12,13,14,15,16], Hosomi-Sakurai allylations [17,18], Michael additions [19], Diels–Alder reactions [20,21,22,23,24,25,26,27,28,29,30,31], halogenations [32], epoxide rearrangements [27], ene reactions [33,34] and other transformations [35,36,37,38]. In these transformations, trityl cations display much higher catalytic reactivity than traditional metal-based Lewis acids. As part of our continuing interests in organosilane chemistry [39,40,41,42], we recently reported (Ph3C)[BPh(F)4]-catalyzed asymmetric Hosomi-Sakurai allylation of chiral crotyl geminal bis(silane) with aldehydes [43]. In this reaction, (Ph3C)[BPh(F)4] proved superior to traditional metal-based Lewis acids. This success led us to extend trityl cation catalysis to allylsilane-mediated reactions for which metal-based Lewis acids do not work well. We focused on β,γ-unsaturated α-ketoesters as electrophiles. Typical enones undergo Michael-type allylation [44,45,46] or [2+2] or [3+2] cyclization [47,48,49,50,51], but Ishihara and co-workers observed that Cu(NTf2)2-catalyzed reaction of β,γ-unsaturated α-ketoesters with allylsilanes gave only the inverse-electron-demand Diels-Alder (IEDDA) reaction adducts as the major products (Scheme 2a)[52]. Sugimura and co-workers achieved the desired allylation using β,γ-unsaturated α,α-dimethoxy esters as the variant (Scheme 2b) [53]. A stoichiometric amount of BF3•Et2O (1.1 equiv.) was required to give γ-substituted α,β-unsaturated α-methoxy esters as a mixture of Z/E isomers.

Scheme 2

Reaction of allylsilane with β,γ-unsaturated α-ketoesters and their variants. “L*” refers to chiral ligands.

Here we report a (Ph3C)[BPh(F)4]-catalyzed Hosomi-Sakurai allylation of allylsilanes with β,γ-unsaturated α-ketoesters (Scheme 2c). The trityl cation shows high catalytic efficiency, giving the γ,γ-disubstituted α-ketoesters in high yields with excellent chemoselectivity. Mechanistic studies suggest that silyl cation catalysis is not a major pathway. Instead, the reaction most likely proceeds via trityl cation catalysis, although we cannot completely rule out Brønsted acid catalysis.

2. Results and Discussion

2.1. Synthesis

We initially screened the metal-based Lewis acid catalysts using β,γ-unsaturated α-ketoester 1a and allytrimethylsilane 2a as model substrates in CH2Cl2 at 25 °C (Table 1). In the presence of 10 mol % of TiCl4, inverse-electron-demand Diels-Alder adduct (±)-4a was obtained in 97% yield; no desired Hosomi-Sakurai allylation product 3a was detected (entry 1). SnCl4, AlCl3 and FeCl3 provided a mixture of 3a and (±)-4a, in which 3a was the minor isomer and the 3a:(±)-4a ratio ranged from 23:77 to 45:55 (entries 2–4). BF3•Et2O and TMSOTf proved to be ineffective catalysts, leading to less than 30% conversion even after 4 days (entries 5 and 6). We also tested lanthanide-based Lewis acids such as Sc(OTf)3 and Yb(OTf)3 (entries 7 and 8). Sc(OTf)3 afforded undesired (±)-4a as the sole detectable product; no reaction occurred using Yb(OTf)3. In sharp contrast to metal-based Lewis acids, the trityl salt (Ph3C)[BPh(F)4] displayed excellent catalytic ability, generating 3a as a single chemoisomer in 97% yield (entry 9). In fact, 1 mol % of (Ph3C)[BPh(F)4] was efficient enough to provide 3a in comparably high yield and selectivity (entry 10).

Table 1

Screening of Reaction Conditions .

Entry	Cat.	Time	3a[(±)-4a] [%] b,c	3a/(±)-4a d
1	TiCl4 (10 mol %)	10 min	0 (97)	≤5:95
2	SnCl4 (10 mol %)	10 min	22 (74)	23:77
3	AlCl3 (10 mol %)	10 min	32 (63)	33:67
4	FeCl3 (10 mol %)	10 min	43 (51)	45:55
5	BF3•Et2O (10 mol %)	4 days	0 (10)	≤5:95
6	TMSOTf (10 mol %)	4 days	20 (<5)	17:83
7	Sc(OTf)3 (10 mol %)	8 h	0 (95)	≤5:95
8	Yb(OTf)3 (10 mol %)	24 h	N.R.	N.D.
9	(Ph3C)[BPh(F)4] (10 mol %)	10 min	97 (0)	≥95:5
10	(Ph3C)[BPh(F)4] (1 mol %)	10 min	97 (0)	≥95:5

Reaction conditions: 0.11 mmol of 1a, 0.13 mmol of 2a in 2.0 mL of CH2Cl2 at 25 °C. The initially formed silyl enol ether was treated with PTS in MeOH to release α-ketoester 3a. Isolated yields. Ratios were determined by 1H NMR spectroscopy of the crude products.

With the optimal reaction conditions in hand, we examined the scope of β,γ-unsaturated α-ketoesters (Scheme 3). Reactions of aryl- substituted ketoesters gave rise to 3b–3j in excellent yields. The reaction tolerated substrates containing functionalized phenyl rings, naphthyl rings or heterocycles. An electron-donating substitution on the phenyl ring slightly decreased chemoselectivity, as shown in 3e (H-S/D-A = 90:10) and 3f (H-S/D-A = 95:5). The reaction generated 3k and 3l from the corresponding alkyl-substituted ketoesters. Allylation of dienyl ketoester with 2a gave 3m in 80% yield, with 1,4-regioselectivity dominating over 1,6-regioselectivity. β,γ-unsaturated α-ketimine ester performed well in the reaction, giving enamido ester 3n in 75% yield. Interestingly, propargyl α-keto ester underwent 1,2-allylation exclusively, leading to α-tertiary hydroxy ester 3o in 93% yield. Switching the ester group from OMe to a bulkier i-Pr group decreased chemoselectivity (3p, H-S/D-A = 91:9).

Scheme 3

Scope of β,γ-Unsaturated α-Ketoesters . Reaction conditions: 0.11 mmol of 1, 0.13 mmol of 2a, 1.0 mol % of (Ph3C)[BPh(F)4] in 2.0 mL of CH2Cl2 at 25 °C. Isolated yields. Ratios were determined by 1H NMR spectroscopy of the crude products.

Trimethylallylsilanes 2b–2e bearing alkyl or aryl substituents at the 2-position reacted well with 1a, giving 3q–3t in 85–96% yields (Table 2, entries 1–4). The high catalytic ability of (Ph3C)[BPh(F)4] also allowed facile anti-SE’ allylation of the bulky 3,3-dimethyl-1-trimethylallylsilanes 2f with 1a (entry 5). However, this catalyst did not efficiently control the diastereoselectivity of allylation: the reaction of Z-crotyltrimethylsilane 2g- afforded 3v in 96% yield but as a 3:2[54] mixture of anti- and syn-diastereomers (entry 6). A similar ratio of 3:1 was obtained using E-crotyltrimethylsilane 2g- (entry 7).

Table 2

Scope of Allylsilanes .

Entry	Allysilanes	Products	Yield
1
2
3
4
5
6
7

Reaction conditions: 0.21 mmol of 1a, 0.26 mmol of 2, 1.0 mol % of (Ph3C)[BPh(F)4] in 4.0 mL of CH2Cl2 at 25 °C. Isolated yields. :syn = 3:2 from 2g-; anti:syn = 3:1 from 2g-.

The trimethylsilyl enol ethers that initially formed in the reactions shown in Scheme 3 and Table 2 were difficult to isolate because of their instability. Switching the silyl moiety from Me3Si to a bulkier Et3Si group in allylsilane 2h–2j led to formation of the stable silyl enol ethers 5a–5c in good to high yields (Scheme 4). The Z-silyl enol ether was favored either as a single isomer (5a and 5c) or the major isomer (5b).

Scheme 4

Formation of Silyl Enol Ethers using Triethylsilylallylsilanes . Reaction conditions: 0.11 mmol of 1a, 0.13 mmol of 2, 1.0 mol % of (Ph3C)[BPh(F)4] in 2.0 mL of CH2Cl2 at 25 °C. Isolated yields. Ratios were determined by 1H NMR spectroscopy of the crude products.

2.2. Mechanistic Investigations

Some mechanistic investigations have been performed for trityl cation catalysis by different groups, but the results appear to be contradictory, particularly in the case of reactions involving allylsilanes or silyl enol ethers. For example, three catalytic species have been suggested for Mukaiyama aldol reactions. Denmark [14] and Mukaiyama [9,10,11,12,13,17,19] proposed the catalytic species to be a trityl cation. In this path, intramolecular transfer of the silyl group releases the product and regenerates the trityl cation catalyst. Bosnich [18] and Chen [16] proposed the catalytic species to be a silyl cation, which is a stronger Lewis acid than trityl cation. In another case, Kagan [20,21,22] proposed the catalytic species to be a Brønsted acid, potentially generated by decomposition of the trityl cation.

The accessibility of silyl enol ethers allowed us to perform detailed mechanistic investigations for our reaction (Scheme 5). Allylsilanes 2a, 2i, 2h and 2b were reacted separately with β,γ-unsaturated α-ketoester 1a. In the merged 1H NMR spectra of the resulting crude silyl enol ethers 5d, 5b, 5a and 5e, we were able to clearly distinguish the Ha signals of the different products (Scheme 5(b1)). Therefore, we reacted a mixture of 2a (1.2 equiv.) and 2i (1.2 equiv.) with 1a (2.0 equiv.) in one pot (Scheme 5a). A mixture of 5d, 5b, 5a and 5e was generated in a ratio of 93(5d + 5b):7(5a + 5e) (Scheme 5(b2)). We attribute the formation of 5a and 5e to crossed silyl cation catalysis. This result implies that 7% of 5d and 5b may form via silyl cation catalysis, meaning that the ratio of trityl to silyl cation catalysis should be approximately (93−7):(7+7) or 86:14. Next we reacted a mixture of 2h (1.2 equiv.) and 2b (1.2 equiv.) with 1a (2.0 equiv.) in one pot. A mixture of 5d, 5b, 5a and 5e was generated in a ratio of 6(5d + 5b):94(5a + 5e) (Scheme 5(b3)). The ratio of trityl to silyl cation catalysis in this reaction should be (94−6):(6+6) or 88:12. The results from these two control reactions suggest that silyl cation catalysis occurs but makes a minor contribution to our results.

Scheme 5

Mechanistic studies (a); 1H NMR spectra of enol ethers 5d, 5b, 5a and 5e (b); possible competing catalytic pathway (c); proposed catalytic cycle (d).

Brønsted acid catalysis is another competing catalytic pathway, which we cannot rule out currently. This pathway seems unlikely to make a major contribution based on our observations (Scheme 5c) that in the presence of 1.0 equiv. of Ph3C+•BF4−, the desired allylation product 3a was obtained in 45% yield, while the by-product (±)-4a also formed in 45% yield. However, neither 3a nor (±)-4a was detected when 1.0 equiv. of HBF4 was used.

Based on these results, we propose a trityl cation-based catalytic mechanism (Scheme 5d). Activation of the ketone in β,γ-unsaturated α-ketoester 1 by trityl cation generates 6 [55]. Subsequent allylation with allylsilane may occur via an unusual closed transition state 7, which allows internal C-to-O silyl transfer to give the non-crossed silyl enol ether 5 as the major product (Scheme 5(b2,b3)). This also regenerates the trityl cation and catalyzes the next cycle.

3. Materials and Methods

3.1. General Procedures for the Synthesis of γ,γ-Disubstituted α-Ketoesters 3a–3v

To a solution of β,γ-unsaturated α-ketoester 1 (0.11 mmol) and allylsilane 2 (0.13 mmol) and in anhyd. CH2Cl2 (2 mL) under argon atmosphere was added (Ph3C)[BPh(F)4] (1.0 mol %) at 25 °C. After stirring for 10 min, the reaction was quenched with p-TsOH (0.5 M in MeOH, 0.1 mL). The mixture was directly concentrated under reduced pressure. Purification of the crude residue via silica gel flash column chromatography (gradient eluent: 0–2.0% of EtOAc/petroleum ether) afforded γ,γ-disubstituted α-ketoester 3. The characterization data for all synthetic compounds are provided in the Supplementary Materials.

Methyl 2-oxo-4-phenylhept-6-enoate (3a): 1H NMR (400 MHz, CDCl3) δ 2.37 (dd, 1H, J1 = 7.2 Hz, J2 = 14.0 Hz), 2.45 (dd, 1H, J1 = 7.2 Hz, J2 = 14.0 Hz), 3.19 (d, 2H, J = 7.2 Hz), 3.34 (dddd, 1H, J = 7.2 Hz), 3.80 (s, 3H), 5.00 (d, 1H, J = 8.0 Hz), 5.03 (d, 1H, J = 15.2 Hz), 5.66 (m, 1H), 7.20 (m, 3H), 7.28 (d, 1H, J = 3.6 Hz), 7.31 (d, 1H, J = 7.6 Hz); 13C NMR (100 MHz, CDCl3) δ 40.2, 40.8, 44.9, 52.9, 117.3, 126.7, 127.5, 128.5, 135.8, 143.2, 161.3, 192.9; IR (neat) cm−1 3029, 2954, 2923, 1731, 1442, 1277, 1253, 1089, 919; HRMS (MALDI, m/z) calcd for C14H16O3Na (M+Na)+: 255.0992, found 255.0989.

3.2. General Procedures for the Synthesis of Silyl Enol Ethers 5a–5e

To a solution of β,γ-unsaturated α-ketoester 1 (0.22 mmol) and allylsilane 2 (0.26 mmol) and in anhyd. CH2Cl2 (4 mL) under argon atmosphere was added (Ph3C)[BPh(F)4] (1.0 mol %) at 25 °C. After stirring for 10 min, the reaction was quenched with NEt3 (1.32 mmol). The mixture was directly concentrated under reduced pressure. Purification of the crude residue via silica gel flash column chromatography (gradient eluent: 0–2.0% of EtOAc/petroleum ether) afforded silyl enol ether 5. The characterization data for all synthetic compounds are provided in the Supplementary Materials.

Methyl (Z)-4-phenyl-2-((triethylsilyl)oxy)hepta-2,6-dienoate (5a): 1H NMR (600 MHz, CDCl3) δ 0.79 (q, 6H, J = 7.8 Hz), 1.05 (t, 9H, J = 7.8 Hz), 2.56 (m, 2H), 3.82 (s, 3H), 3.99 (m, 1H), 5.05 (d, 1H, J = 10.2 Hz), 5.11 (d, 1H, J = 16.8 Hz), 5.78 (m, 1H), 7.30 (m, 3H), 7.38 (m, 2H); 13C NMR (100 MHz, CDCl3) δ 5.5, 6.8, 40.6, 41.8, 51.9, 116.4, 124.6, 126.4, 127.5, 128.5, 135.0, 140.1, 143.3, 165.3; IR (neat) cm−1 2955, 2878, 1727, 1642, 1439, 1372, 1266, 1232, 1143, 1010, 913; HRMS (MALDI, m/z) calcd for C20H30O3SiNa (M+Na)+: 369.1856, found 369.1856.

4. Conclusions

In summary, we have developed a (Ph3C)[BPh(F)4]-catalyzed Hosomi-Sakurai allylation of allylsilanes with β,γ-unsaturated α-ketoesters. Various γ,γ-disubstituted α-ketoesters α-ketoesters were synthesized in high yields with excellent chemoselectivity. Mechanistic studies suggest that the trityl cation dominates catalysis, while the silyl cation plays only a minor role. Verification of this mechanism also makes the trityl cation-catalyzed asymmetric reaction possible, which is a challenging task and little progress has been achieved. The related work is ongoing in our group.