Remote Tris(pentafluorophenyl)borane-Assisted Chiral Phosphoric Acid Catalysts for the Enantioselective Diels–Alder Reaction

Abstract

Tris(pentafluorophenyl)borane-assisted chiral supramolecular phosphoric acid catalysts were developed for the model Diels–Alder reaction of α-substituted acroleins with cyclopentadiene. Two remotely coordinated tris(pentafluorophenyl)boranes should help to increase the Brønsted acidity of the active center in the supramolecular catalyst and create effective bulkiness for the chiral cavity. The prepared supramolecular catalysts acted as not only conjugated Brønsted acid–Brønsted base catalysts but also bifunctional Lewis acid–Brønsted base catalysts with the addition of a central achiral Lewis acid source such as catecholborane.

The design of simple artificial enzymes is an ongoing challenge in modern organic chemistry. In particular, tailor-made chiral supramolecular catalysts might be attractive for use as artificial enzymes, since every part of a supramolecular catalyst can be fine-tuned for each substrate to establish higher-ordered substrate-selectivity and/or stereoselectivity.[1] [2] In this regard, we previously developed enantioselective Diels–Alder reactions with anomalous endo/exo-selectivities through the use of conformationally flexible chiral supramolecular Lewis acid catalyst 1 (Figure [1, a]).[3] [4]

Based on the chiral deep and narrow cavity control of the transition states in the reaction of α-substituted acroleins and cyclopentadiene, anomalous endo products were successfully obtained in a highly enantioenriched fashion for the first time.[5] Two coordinated tris(pentafluorophenyl)boranes in catalyst 1 should not only create effective bulkiness for the chiral cavity but also help to increase the Lewis acidity of the active center based on the Lewis acid assisted Lewis acid (LLA)[6] catalyst system. To further develop such a supramolecular methodology, we envisioned that we might be able to use conformationally flexible chiral supramolecular Brønsted acid catalyst 2 based on the Lewis acid assisted Brønsted acid (LBA)[6] catalyst system (Figure [1, b]). By taking advantage of the conjugated Brønsted acid–Brønsted base bifunction of chiral phosphoric acids 2,[7] [8] aldehydes (i.e., acroleins) should be able to doubly coordinate with the active centers (Figure [2, a]). Moreover, the addition of an achiral Lewis acid source (ML _n ) should provide bifunctional Lewis acid–Brønsted base catalysts (Figure [2, b]). Overall, introduction of the phosphoric acid to the center of supramolecular catalysts might provide additional opportunities for versatile molecular recognition according to the size and/or substitution pattern of acroleins. In this context, we have developed remote tris(pentafluorophenyl)borane-assisted chiral phosphoric acid catalysts for the enantioselective Diels–Alder reaction of α-substituted acroleins with cyclopentadiene as a probe reaction.

First, we examined the Diels–Alder reaction of meth­acrolein 5a with cyclopentadiene 4 in dichloromethane at –78 °C in the presence of chiral supramolecular catalysts (10 mol%), which were prepared in situ from chiral phosphoric acid (R)-3a and achiral boron Lewis acids, such as BF₃·Et₂O, BBr₃, and B(C₆F₅)₃ (Table [1]). The reaction did not proceed with the use of (R)-3a alone (Table [1], entry 1). In contrast, the combined use of (R)-3a and Lewis acids showed strong catalytic activities (Table [1], entries 2–4). In particular, as expected, bulky B(C₆F₅)₃ was more effective than BF₃·Et₂O and BBr₃, and higher enantioselectivity (53% ee) was observed (Table [1], entry 4). Fortunately, the enantio­selectivity was significantly improved when amide-type (R)-3b and (R)-3c were used in place of phosphoryl-type (R)-3a, and exo-6a was obtained with 90% ee (Table [1], entries 10 and 16). Moreover, for (R)-3b and (R)-3c as well as (R)-3a, BF₃·Et₂O and BBr₃ were not effective (Table [1], entries 8, 9, 14, and 15).

In this reaction, preparation of the catalyst at room temperature in advance was critical,[9] and compounds 4 and 5a were added within five minutes just after the mixture of catalysts was cooled to –78 °C.[10] In this regard, the enantio­selectivity significantly decreased when compounds 4 and 5a were added to the mixture of 2B(C₆F₅)₃–(R)-3a or 2B(C₆F₅)₃–(R)-3b after cooling at –78 °C for 30 minutes (Table [1], entries 5 and 11). Once B(C₆F₅)₃ is adventitiously released from the supramolecular catalysts 2B(C₆F₅)₃–(R)-3a and 2B(C₆F₅)₃–(R)-3b, the highly basic phosphoryl moiety and pyrrolidine-derived amido moiety would tightly coordinate with the proton of the phosphoric acid at –78 °C. The corresponding species B(C₆F₅)₃–(R)-3a and B(C₆F₅)₃–(R)-3b, which might be inactive (Table [1], entries 6 and 12), would then be formed (Figure [3]). Simultaneously, achiral B(C₆F₅)₃, which might induce a racemic reaction pathway (Table [1], entries 5 and 11), would be released. In sharp contrast, 2B(C₆F₅)₃–(R)-3c, which has a much less basic isoindoline-derived amido moiety, was tolerated under the reaction conditions at –78 °C for 30 minutes before the addition of substrates 4 and 5a, and exo-6a was obtained with 85% ee (Table [1], entry 17). The inter-/intramolecular coordination-exchange between the proton center and B(C₆F₅)₃ might still occur due to the weak basicity of the isoindoline-derived amido moiety even at –78 °C. As a result, active 2B(C₆F₅)₃–(R)-3c would be regenerated from less active B(C₆F₅)₃–(R)-3c. These considerations might be supported by finding that the catalytic activity of 2B(C₆F₅)₃–(R)-3c (entry 16) was much higher than those of competitive B(C₆F₅)₃–(R)-3c (Table [1], entry 18) and free B(C₆F₅)₃ (Table [1], entry 19).[11]

Table 1 Screening of Chiral Supramolecular Catalysts^a

Entry	(R)-3	Lewis acid	Yield (%)	endo-6a/exo-6a	ee (%) of exo-6a
1	(R)-3a	–	0	–	–
2	(R)-3a	BF3·Et2O	>99	7:93	2
3	(R)-3a	BBr3	98	8:92	14
4	(R)-3a	B(C6F5)3	98	8:92	–53b
5c	(R)-3a	B(C6F5)3	>99	8:92	–31b
6d	(R)-3a	B(C6F5)3	20	19:81	–1b
7	(R)-3b	–	0	–	–
8	(R)-3b	BF3·Et2O	>99	7:93	0
9	(R)-3b	BBr3	>99	5:95	0
10	(R)-3b	B(C6F5)3	>99	5:95	90
11c	(R)-3b	B(C6F5)3	62	11:89	8
12d	(R)-3b	B(C6F5)3	0	–	–
13	(R)-3c	–	0	–	–
14	(R)-3c	BF3·Et2O	98	6:94	0
15	(R)-3c	BBr3	96	10:90	0
16	(R)-3c	B(C6F5)3	>99	8:92	90
17c	(R)-3c	B(C6F5)3	>99	9:91	85
18d	(R)-3c	B(C6F5)3	0	–	–
19e	–	B(C6F5)3	>99	7:93	–

^a Unless otherwise noted, the reaction of 5a (0.5 mmol) with 4 (2.5 mmol) was carried out with the use of (R)-3 (10 mol%), Lewis acid (20 mol%), and MS 4Å in CH₂Cl₂ at –78 °C for 1 h. Compounds 4 and 5a were added to the mixture of catalysts within 5 min just after cooling to –78 °C.

^b Enantioselectivity of exo-(2R)-6a.

^c Compounds 4 and 5a were added to the mixture of catalysts after it was cooled to –78 °C for 30 min.

^d The reaction was carried out with the use of (R)-3 (10 mol%), B(C₆F₅)₃ (10 mol%), and MS 4Å under standard conditions.

^e The reaction was carried out with the use of B(C₆F₅)₃ (20 mol%) alone.

To confirm the complexation of the optimized supramolecular catalyst 2B(C₆F₅)₃–(R)-3c, spectroscopic analyses were performed at room temperature (Scheme [1]). We found a peak at 1697.146 in ESI–MS analysis (negative mode), which might be unambiguously attributed to {[2B(C₆F₅)₃–(R)-3c] + 2H₂O – H}^– (see the Supporting Information). Moreover, peaks at δ = –137.0, –159.5, and –166.3 ppm in ¹⁹F NMR (CD₂Cl₂) were slightly shifted from the original peaks of B(C₆F₅)₃ at δ = –130.2, –147.1, and –161.4 ppm. However, a peak at δ = 4.6 ppm in ³¹P NMR (CD₂Cl₂) was scarcely shifted from the original peak at δ = +4.0 ppm. These observations suggest that coordination to B(C₆F₅)₃ at the carbonyl groups of the 3,3′-substituents would precede coordination at the central P=O moiety, probably due to steric constraints at the narrow inner space.[12] Unfortunately, the hydrogen-bonding structures of B(C₆F₅)₃–(R)-3a and B(C₆F₅)₃–(R)-3b, as shown in Figure [3] have not yet been confirmed directly (see the Supporting Information for ³¹P NMR analysis.). However, in ³¹P NMR (CD₂Cl₂) analysis at –78 °C, a peak at δ = –4.1 ppm of 2B(C₆F₅)₃–(R)-3b gradually decreased, and many other peaks between δ = –5 ppm and δ = –25 ppm were predominantly observed within 30 minutes.[13] In sharp contrast, the decomposition of 2B(C₆F₅)₃–(R)-3c to other species at –78 °C was much slower than that of 2B(C₆F₅)₃–(R)-3b, and ca. 70% of 2B(C₆F₅)₃–(R)-3b was still observed at δ = –0.7 ppm for at least 30 minutes.[13]

Next, we examined the substrate specificity for α-substituted acroleins. In place of methacrolein 5a, α-ethylacrolein 5b could be used in the presence of 2B(C₆F₅)₃–(R)-3c, and the corresponding normal exo-6b was obtained with 84% ee (Table [2], entry 3). Partially due to steric mismatch with the chiral cavity, α-isopropylacrolein 5c and α-bromoacrolein 5d, which are bulkier than 5a and 5b, might be unsuitable for the chiral cavity of 2B(C₆F₅)₃–(R)-3c, and low enantioselectivities were observed (Table [2], entries 5 and 7). Moreover, a racemic pathway also might be promoted in the case of highly reactive 5d (Table [2], entry 7). As compared with thermal conditions (Table [2], entries 2, 4, 6, and 8), a supramolecular catalyst induced a slight exo preference for 6a and 6b (Table [2], entries 1 and 3).

Moreover, with the use of α-nonsubstituted acrolein 5e, moderate anomalous exo-selectivity was observed (endo/exo = 49:51) (Scheme [2]).[14] Although the enantioselectivities of endo-6e and exo-6e were low (30% ee and 25% ee, respectively), an unusual disagreement in stereoselectivity (R/S) was observed between normal endo-(2S)-6e and anomalous exo-(2R)-6e. These results suggest that 2B(C₆F₅)₃–(R)-3c might have an exo-inducing chiral cavity as a supramolecular catalyst.

Table 2 Substrate Specificity with the Use of 2B(C₆F₅)₃–(R)-3c ^a

Entry	5 (R)	Product	Yield (%)	endo-6/exo-6	ee (%) of exo-6
1	5a (Me)	6a	>99	8:92	90 (2S)
2b	5a (Me)	6a	94 (40 °C, 3 h)	16:84	–
3	5b (Et)	6b	>99	2:98	84 (2S)
4b	5b (Et)	6b	73 (110 °C, 24 h)	24:76	–
5	5c (i-Pr)	6c	72	15:85	23 (2R)
6b	5c (i-Pr)	6c	<5 (110 °C, 3 h)	–	–
7	5d (Br)	6d	>99	15:85	18 (2R)
8b	5d (Br)	6d	>99 (r.t., 3 h)	15:85	–

^a The reaction of 5 (0.5 mmol) with 4 (2.5 mmol) was carried out with the use of (R)-3c (10 mol%), B(C₆F₅)₃ (20 mol%), and MS 4Å in CH₂Cl₂ at –78 °C for 1 h.

^b The reactions were carried out under thermal conditions without any catalysts in CH₂Cl₂ at room temperature to 40 °C or in toluene at 110 °C.

Finally, we used tiglic aldehyde 7 as a much less reactive α,β-disubstituted acrolein, which did not give product 8 under thermal conditions in toluene at 110 °C. Supramolecular catalyst 2B(C₆F₅)₃–(R)-3c showed low reactivity, and exo-8 was obtained in 38% yield with 56% ee (Scheme [3, a]). To improve both the yield and the enantioselectivity, we changed the Brønsted acid–Brønsted base catalyst system (Figure [2, a]) to a Lewis acid–Brønsted base catalyst system (Figure [2, b]), by using an additional achiral Lewis acid partner. After screening the acid sources,[15] we found that catecholborane was highly effective as a boron Lewis acid center for 2B(C₆F₅)₃–(R)-3c, and exo-8 was obtained in 71% yield with 75% ee (Scheme [3, b]).[16] Although the enantioselectivity has still been moderate, these results represent at least a partial demonstration of our conceptual catalytic system in Figure [2].

In this preliminary stage, further information based on experimental and theoretical studies will be necessary to discuss possible structures of the supramolecular catalysts in situ. In this regard, the previous supramolecular catalyst 1 has been calculated to be the C ₁-symmetric syn conformation due to the sp ³ boron Lewis acid center.[3] In contrast, we can speculate that supramolecular catalyst 2B(C₆F₅)₃–(R)-3c would have an anti conformation as shown in Figure [4] (b), unlike a sterically hindered syn conformation as shown in Figure [4] (a), due to the essentially C ₂-symmetric conjugated phosphoric acid moiety. Catecholborane-introduced supramolecular catalyst might have similar structures although the field would then be C ₁-symmetric, as shown in Figure [4] (c). In anti conformations, as shown in Figure [4] (b and c), a shallow and wide chiral cavity would be formed around the active center which would induce substrate specificity with an exo-preference.[3b]

In summary, we have developed bulky and strong Lewis acid B(C₆F₅)₃-assisted chiral phosphoric acids, which were designed for the model Diels–Alder reaction of α-substituted acroleins with cyclopentadiene.[17] The corresponding supramolecular catalysts acted not only as highly activated conjugated Brønsted acid–Brønsted base catalysts but also as bifunctional Lewis acid–Brønsted base catalysts with the addition of a central achiral Lewis acid source such as catecholborane. Further investigations with these asymmetric supramolecular methodologies with the use of chiral phosphoric acids, which might contribute to the construction of a conformationally flexible, bulky, and chiral cavity for higher-ordered catalysis, are currently underway.[18]

Acknowledgment

Financial support was partially provided by JSPS, KAKENHI (15H05755, 26288046, and 26105723), Program for Leading Graduate Schools ‘IGER program in Green Natural Sciences’, MEXT, Japan.

• References and Notes

• 1 Lehn JM. Science 1985; 227: 849
  CrossrefSearch in Google Scholar

For recent reviews on supramolecular catalysis, see:
• 2a Ma L, Abney C, Lin W. Chem. Soc. Rev. 2009; 38: 1248
  CrossrefSearch in Google Scholar
• 2b Corma A, García H, Llabrés i Xamena FX. Chem. Rev. 2010; 110: 4606
  CrossrefSearch in Google Scholar
• 2c Meeuwissen J, Reek JN. H. Nature Chem. 2010; 2: 615
  CrossrefSearch in Google Scholar
• 2d Raynal M, Ballester P, Vidal-Ferran A, van Leeuwena PW. N. M. Chem. Soc. Rev. 2014; 43: 1660
  CrossrefSearch in Google Scholar
• 2e Raynal M, Ballester P, Vidal-Ferran A, van Leeuwena PW. N. M. Chem. Soc. Rev. 2014; 43: 1734
  CrossrefSearch in Google Scholar
• 2f Brown CJ, Toste FD, Bergman RG, Raymond KN. Chem. Rev. 2015; 115: 3012
  CrossrefSearch in Google Scholar
• 3a Hatano M, Mizuno T, Izumiseki A, Usami R, Asai T, Akakura M, Ishihara K. Angew. Chem. Int. Ed. 2011; 50: 12189
  CrossrefPubMedSearch in Google Scholar
• 3b Hatano M, Ishihara K. Chem. Commun. 2012; 48: 4273
  CrossrefSearch in Google Scholar

For reviews on the Diels–Alder reaction, see:
• 4a Kagan HB, Riant O. Chem. Rev. 1992; 92: 1007
  Search in Google Scholar
• 4b Du H, Ding K In Handbook of Cyclization Reactions . Ma S. Wiley-VCH; Stuttgart: 2010: 1-57
  Search in Google Scholar
• 4c Ishihara K, Sakakura A In Science of Synthesis, Stereoselective Synthesis . Vol. 3. Evans PA. Thieme; Stuttgart: 2011: 67-123
  Search in Google Scholar

In general, endo/exo-selectivity in the Diels–Alder reaction depends on the substrates, see:
• 5a Hoffmann R, Woodward RB. J. Am. Chem. Soc. 1965; 87: 4388
  CrossrefSearch in Google Scholar
• 5b García JI, Mayoral JA, Salvatella L. Acc. Chem. Res. 2000; 33: 658
  CrossrefSearch in Google Scholar
• 5c Barba C, Carmona D, García JI, Lamata MP, Mayoral JA, Salvatella L, Viguri F. J. Org. Chem. 2006; 71: 9831
  CrossrefSearch in Google Scholar
• 5d Wannere CS, Paul A, Herges R, Houk KN, Schaefer HF. III, von Ragué Schleyer P. J. Comput. Chem. 2007; 28: 344
  CrossrefSearch in Google Scholar

Yamamoto developed the pioneering concept of combined acid catalysis. See reviews:
• 6a Ishibashi H, Ishihara K, Yamamoto H. Chem. Rec. 2002; 2: 177
  CrossrefSearch in Google Scholar
• 6b Yamamoto H, Futatsugi K. Angew. Chem. Int. Ed. 2005; 44: 1924
  CrossrefPubMedSearch in Google Scholar
• 6c Yamamoto H, Futatsugi K In Acid Catalysis in Modern Organic Synthesis . Vol. 1. Yamamoto H, Ishihara K. Wiley-VCH; Weinheim: 2008: 1-34
  Search in Google Scholar

For reviews on chiral phosphoric acids, see:
• 7a Akiyama T. Chem. Rev. 2007; 107: 5744
  CrossrefPubMedSearch in Google Scholar
• 7b Terada M. Synthesis 2010; 1929
  Thieme Connect
• 7c Parmar D, Sugiono E, Raja S, Rueping M. Chem. Rev. 2014; 114: 9047
  CrossrefPubMedSearch in Google Scholar

Reviews on acid–base combination chemistry, see:
• 8a Kanai M, Kato N, Ichikawa E, Shibasaki M. Synlett 2005; 1491
  Thieme Connect
• 8b Ishihara K, Sakakura A, Hatano M. Synlett 2007; 686
• 8c Ishihara K. Proc. Jpn. Acad. Ser. B 2009; 85: 290
  CrossrefSearch in Google Scholar
• 8d Shibasaki M, Kanai M, Matsunaga S, Kumagai N. Acc. Chem. Res. 2009; 42: 1117
  CrossrefPubMedSearch in Google Scholar
• 9 For 2B(C₆F₅)₃–(R)-3a/3b/3c, the aging time (0.5–2 h) at r.t. in advance showed no significant differences in the Diels–Alder reaction.
• 10 Lower enantioselectivities (ca. 60% ee) were observed when a solution of the catalyst 2B(C₆F₅)₃–(R)-3b or 2B(C₆F₅)₃–(R)-3c at r.t. was added to the solution of 4 and 5a at –78 °C.
• 11 Compound 5a is too reactive to evaluate meaningful differences in the catalytic activity between 2B(C₆F₅)₃–(R)-3c and free B(C₆F₅)₃. However, the catalytic activity of 2B(C₆F₅)₃–(R)-3c was much higher than that of free B(C₆F₅)₃. See Scheme 3 and the SI.
• 12 To confirm whether or not the coordination of the P=O moiety to B(C₆F₅)₃ would occur, we used (R)-3,3′-Ph₂BINOL-derived phosphoric acid, which may avoid competitive coordinations. In ³¹P NMR (CD₂Cl₂) analysis at r.t., a singlet peak at δ = +1.7 ppm changed to δ = –1.0 ppm with a small upfield shift, which suggests the coordination of the P=O moiety to B(C₆F₅)₃. Next, as with 2B(C₆F₅)₃–(R)-3c, almost the same shifted peaks at δ = –137.0, –158.8, and –165.8 ppm were observed in ¹⁹F NMR (CD₂Cl₂) at r.t.
• 13 ³¹P NMR (CD₂Cl₂) analysis of (R)-3b and (R)-3c at –78 °C showed a peak at δ = 6.1 ppm and 5.0 ppm, respectively, although solubility of them at –78 °C was low. See the SI.

For anomalous exo-selective Diels–Alder reactions of α-nonsubstituted acroleins, see:
• 14a Maruoka K, Imoto H, Yamamoto H. J. Am. Chem. Soc. 1994; 116: 12115
  CrossrefSearch in Google Scholar
• 14b Kündig EP, Saudan CM, Alezra V, Viton F, Bernardinelli G. Angew. Chem. Int. Ed. 2001; 40: 4481
  CrossrefSearch in Google Scholar
• 14c Kano T, Tanaka Y, Maruoka K. Org. Lett. 2006; 8: 2687
  CrossrefSearch in Google Scholar
• 14d Hayashi Y, Samanta S, Gotoh H, Ishikawa H. Angew. Chem. Int. Ed. 2008; 47: 6634. Also see ref 3
  CrossrefSearch in Google Scholar
• 15 We examined Me₃Al, Et₃Al, i-Bu₂AlH (DIBAL-H), Me₂AlNTf₂, allyltrimethylsilane, pinacolborane, 9-borabicyclo[3.3.1]nonane (9-BBN), etc. However, the combined use of these achiral Lewis acid sources to 2B(C₆F₅)₃–(R)-3c showed low reactivities (0–15% yields), and the sole exception was catecholborane. In this regard, the combined use of a stoichiometric amount of catecholborane with chiral phosphoric acid catalyst in the enantio­selective reduction of ketones was reported by Antilla. See: Zhang Z, Jain P, Antilla JC. Angew. Chem. Int. Ed. 2011; 50: 10961
  CrossrefSearch in Google Scholar
• 16 We examined the reactions of acroleins 5a–e with cyclopentadiene 4 with the use of a supramolecular catalyst, which was prepared from (R)-3c, B(C₆F₅)₃, and catecholborane, However, better enantioselectivities were not observed compared with 2B(C₆F₅)₃–(R)-3c as shown in Table 2 and Scheme 2. The results are summarized in the SI.
• 17 Typical Procedure for the Diels–Alder Reaction: To a mixture of (R)-3c (31.9 mg, 0.050 mmol) and powdered MS 4Å (200 mg) in a Schlenk tube under a nitrogen atmosphere, tris(pentafluorophenyl)borane (51.2 mg, 0.10 mmol) and freshly distilled CH₂Cl₂ (2 mL) were added via a cannula, and this suspension was stirred at r.t. for 1 h. Next, the mixture was cooled to –78 °C, and as soon as possible (within 5 min) after cooling to –78 °C, methacrolein 5a (95% purity, 43.4 μL, 0.50 mmol) and freshly distilled cyclopentadiene 4 (203 μL, 2.5 mmol) were added at –78 °C. After that, the resultant mixture was stirred at –78 °C for 1 h. To quench the reaction, Et₃N (0.2 mL) was poured into the reaction mixture at –78 °C. The product mixture was warmed to r.t. and directly purified by silica gel column chromatography (eluent: pentane–Et₂O, 9:1). Solvents were removed under 200 Torr at 20 °C by a rotary evaporator, and the product 6a was obtained (68.2 mg, >99% yield). ¹H NMR (400 MHz, CDCl₃): δ = 0.76 (d, J = 12.0 Hz, 1 H), 1.01 (s, 3 H), 1.39 (m, 2 H), 2.25 (dd, J = 12.0, 3.9 Hz, 1 H), 2.82 (br s, 1 H), 2.90 (br s, 1 H), 6.11 (dd, J = 6.0, 3.0 Hz, 1 H), 6.30 (dd, J = 6.0, 3.0 Hz, 1 H), 9.69 (s, 1 H). ¹³C NMR (100 MHz, CDCl₃): δ = 20.1, 34.6, 43.2, 47.6, 48.5, 53.9, 133.1, 139.6, 205.9. HRMS (EI): m/z [M]⁺ calcd for C₉H₁₂O: 136.0888; found: 136.0893. The endo/exo ratio of 6a was determined by NMR analysis. ¹H NMR (CDCl₃): δ = 9.40 [s, 1 H, CHO (endo-6a)], 9.69 [s, 1 H, CHO (exo-6a)]; see ref 3a. The enantioselectivity and absolute stereochemistry of 6a were determined by GC analysis according to the literature (see ref. 3a).
• 18 We just recently reported boron tribromide assisted chiral phosphoric acid catalyst for a highly enantioselective Diels–Alder reaction of 1,2-dihydropyridines. See: Hatano M, Goto Y, Izumiseki A, Akakura M, Ishihara K. J. Am. Chem. Soc. 2015; 137: 13472
  CrossrefSearch in Google Scholar

• References and Notes

• 1 Lehn JM. Science 1985; 227: 849
  CrossrefSearch in Google Scholar

For recent reviews on supramolecular catalysis, see:
• 2a Ma L, Abney C, Lin W. Chem. Soc. Rev. 2009; 38: 1248
  CrossrefSearch in Google Scholar
• 2b Corma A, García H, Llabrés i Xamena FX. Chem. Rev. 2010; 110: 4606
  CrossrefSearch in Google Scholar
• 2c Meeuwissen J, Reek JN. H. Nature Chem. 2010; 2: 615
  CrossrefSearch in Google Scholar
• 2d Raynal M, Ballester P, Vidal-Ferran A, van Leeuwena PW. N. M. Chem. Soc. Rev. 2014; 43: 1660
  CrossrefSearch in Google Scholar
• 2e Raynal M, Ballester P, Vidal-Ferran A, van Leeuwena PW. N. M. Chem. Soc. Rev. 2014; 43: 1734
  CrossrefSearch in Google Scholar
• 2f Brown CJ, Toste FD, Bergman RG, Raymond KN. Chem. Rev. 2015; 115: 3012
  CrossrefSearch in Google Scholar
• 3a Hatano M, Mizuno T, Izumiseki A, Usami R, Asai T, Akakura M, Ishihara K. Angew. Chem. Int. Ed. 2011; 50: 12189
  CrossrefPubMedSearch in Google Scholar
• 3b Hatano M, Ishihara K. Chem. Commun. 2012; 48: 4273
  CrossrefSearch in Google Scholar

For reviews on the Diels–Alder reaction, see:
• 4a Kagan HB, Riant O. Chem. Rev. 1992; 92: 1007
  Search in Google Scholar
• 4b Du H, Ding K In Handbook of Cyclization Reactions . Ma S. Wiley-VCH; Stuttgart: 2010: 1-57
  Search in Google Scholar
• 4c Ishihara K, Sakakura A In Science of Synthesis, Stereoselective Synthesis . Vol. 3. Evans PA. Thieme; Stuttgart: 2011: 67-123
  Search in Google Scholar

In general, endo/exo-selectivity in the Diels–Alder reaction depends on the substrates, see:
• 5a Hoffmann R, Woodward RB. J. Am. Chem. Soc. 1965; 87: 4388
  CrossrefSearch in Google Scholar
• 5b García JI, Mayoral JA, Salvatella L. Acc. Chem. Res. 2000; 33: 658
  CrossrefSearch in Google Scholar
• 5c Barba C, Carmona D, García JI, Lamata MP, Mayoral JA, Salvatella L, Viguri F. J. Org. Chem. 2006; 71: 9831
  CrossrefSearch in Google Scholar
• 5d Wannere CS, Paul A, Herges R, Houk KN, Schaefer HF. III, von Ragué Schleyer P. J. Comput. Chem. 2007; 28: 344
  CrossrefSearch in Google Scholar

Yamamoto developed the pioneering concept of combined acid catalysis. See reviews:
• 6a Ishibashi H, Ishihara K, Yamamoto H. Chem. Rec. 2002; 2: 177
  CrossrefSearch in Google Scholar
• 6b Yamamoto H, Futatsugi K. Angew. Chem. Int. Ed. 2005; 44: 1924
  CrossrefPubMedSearch in Google Scholar
• 6c Yamamoto H, Futatsugi K In Acid Catalysis in Modern Organic Synthesis . Vol. 1. Yamamoto H, Ishihara K. Wiley-VCH; Weinheim: 2008: 1-34
  Search in Google Scholar

For reviews on chiral phosphoric acids, see:
• 7a Akiyama T. Chem. Rev. 2007; 107: 5744
  CrossrefPubMedSearch in Google Scholar
• 7b Terada M. Synthesis 2010; 1929
  Thieme Connect
• 7c Parmar D, Sugiono E, Raja S, Rueping M. Chem. Rev. 2014; 114: 9047
  CrossrefPubMedSearch in Google Scholar

Reviews on acid–base combination chemistry, see:
• 8a Kanai M, Kato N, Ichikawa E, Shibasaki M. Synlett 2005; 1491
  Thieme Connect
• 8b Ishihara K, Sakakura A, Hatano M. Synlett 2007; 686
• 8c Ishihara K. Proc. Jpn. Acad. Ser. B 2009; 85: 290
  CrossrefSearch in Google Scholar
• 8d Shibasaki M, Kanai M, Matsunaga S, Kumagai N. Acc. Chem. Res. 2009; 42: 1117
  CrossrefPubMedSearch in Google Scholar
• 9 For 2B(C₆F₅)₃–(R)-3a/3b/3c, the aging time (0.5–2 h) at r.t. in advance showed no significant differences in the Diels–Alder reaction.
• 10 Lower enantioselectivities (ca. 60% ee) were observed when a solution of the catalyst 2B(C₆F₅)₃–(R)-3b or 2B(C₆F₅)₃–(R)-3c at r.t. was added to the solution of 4 and 5a at –78 °C.
• 11 Compound 5a is too reactive to evaluate meaningful differences in the catalytic activity between 2B(C₆F₅)₃–(R)-3c and free B(C₆F₅)₃. However, the catalytic activity of 2B(C₆F₅)₃–(R)-3c was much higher than that of free B(C₆F₅)₃. See Scheme 3 and the SI.
• 12 To confirm whether or not the coordination of the P=O moiety to B(C₆F₅)₃ would occur, we used (R)-3,3′-Ph₂BINOL-derived phosphoric acid, which may avoid competitive coordinations. In ³¹P NMR (CD₂Cl₂) analysis at r.t., a singlet peak at δ = +1.7 ppm changed to δ = –1.0 ppm with a small upfield shift, which suggests the coordination of the P=O moiety to B(C₆F₅)₃. Next, as with 2B(C₆F₅)₃–(R)-3c, almost the same shifted peaks at δ = –137.0, –158.8, and –165.8 ppm were observed in ¹⁹F NMR (CD₂Cl₂) at r.t.
• 13 ³¹P NMR (CD₂Cl₂) analysis of (R)-3b and (R)-3c at –78 °C showed a peak at δ = 6.1 ppm and 5.0 ppm, respectively, although solubility of them at –78 °C was low. See the SI.

For anomalous exo-selective Diels–Alder reactions of α-nonsubstituted acroleins, see:
• 14a Maruoka K, Imoto H, Yamamoto H. J. Am. Chem. Soc. 1994; 116: 12115
  CrossrefSearch in Google Scholar
• 14b Kündig EP, Saudan CM, Alezra V, Viton F, Bernardinelli G. Angew. Chem. Int. Ed. 2001; 40: 4481
  CrossrefSearch in Google Scholar
• 14c Kano T, Tanaka Y, Maruoka K. Org. Lett. 2006; 8: 2687
  CrossrefSearch in Google Scholar
• 14d Hayashi Y, Samanta S, Gotoh H, Ishikawa H. Angew. Chem. Int. Ed. 2008; 47: 6634. Also see ref 3
  CrossrefSearch in Google Scholar
• 15 We examined Me₃Al, Et₃Al, i-Bu₂AlH (DIBAL-H), Me₂AlNTf₂, allyltrimethylsilane, pinacolborane, 9-borabicyclo[3.3.1]nonane (9-BBN), etc. However, the combined use of these achiral Lewis acid sources to 2B(C₆F₅)₃–(R)-3c showed low reactivities (0–15% yields), and the sole exception was catecholborane. In this regard, the combined use of a stoichiometric amount of catecholborane with chiral phosphoric acid catalyst in the enantio­selective reduction of ketones was reported by Antilla. See: Zhang Z, Jain P, Antilla JC. Angew. Chem. Int. Ed. 2011; 50: 10961
  CrossrefSearch in Google Scholar
• 16 We examined the reactions of acroleins 5a–e with cyclopentadiene 4 with the use of a supramolecular catalyst, which was prepared from (R)-3c, B(C₆F₅)₃, and catecholborane, However, better enantioselectivities were not observed compared with 2B(C₆F₅)₃–(R)-3c as shown in Table 2 and Scheme 2. The results are summarized in the SI.
• 17 Typical Procedure for the Diels–Alder Reaction: To a mixture of (R)-3c (31.9 mg, 0.050 mmol) and powdered MS 4Å (200 mg) in a Schlenk tube under a nitrogen atmosphere, tris(pentafluorophenyl)borane (51.2 mg, 0.10 mmol) and freshly distilled CH₂Cl₂ (2 mL) were added via a cannula, and this suspension was stirred at r.t. for 1 h. Next, the mixture was cooled to –78 °C, and as soon as possible (within 5 min) after cooling to –78 °C, methacrolein 5a (95% purity, 43.4 μL, 0.50 mmol) and freshly distilled cyclopentadiene 4 (203 μL, 2.5 mmol) were added at –78 °C. After that, the resultant mixture was stirred at –78 °C for 1 h. To quench the reaction, Et₃N (0.2 mL) was poured into the reaction mixture at –78 °C. The product mixture was warmed to r.t. and directly purified by silica gel column chromatography (eluent: pentane–Et₂O, 9:1). Solvents were removed under 200 Torr at 20 °C by a rotary evaporator, and the product 6a was obtained (68.2 mg, >99% yield). ¹H NMR (400 MHz, CDCl₃): δ = 0.76 (d, J = 12.0 Hz, 1 H), 1.01 (s, 3 H), 1.39 (m, 2 H), 2.25 (dd, J = 12.0, 3.9 Hz, 1 H), 2.82 (br s, 1 H), 2.90 (br s, 1 H), 6.11 (dd, J = 6.0, 3.0 Hz, 1 H), 6.30 (dd, J = 6.0, 3.0 Hz, 1 H), 9.69 (s, 1 H). ¹³C NMR (100 MHz, CDCl₃): δ = 20.1, 34.6, 43.2, 47.6, 48.5, 53.9, 133.1, 139.6, 205.9. HRMS (EI): m/z [M]⁺ calcd for C₉H₁₂O: 136.0888; found: 136.0893. The endo/exo ratio of 6a was determined by NMR analysis. ¹H NMR (CDCl₃): δ = 9.40 [s, 1 H, CHO (endo-6a)], 9.69 [s, 1 H, CHO (exo-6a)]; see ref 3a. The enantioselectivity and absolute stereochemistry of 6a were determined by GC analysis according to the literature (see ref. 3a).
• 18 We just recently reported boron tribromide assisted chiral phosphoric acid catalyst for a highly enantioselective Diels–Alder reaction of 1,2-dihydropyridines. See: Hatano M, Goto Y, Izumiseki A, Akakura M, Ishihara K. J. Am. Chem. Soc. 2015; 137: 13472
  CrossrefSearch in Google Scholar

• Supplementary Material