|Home | About | Journals | Submit | Contact Us | Français|
The development of enantiomerically-resolved, axially-chiral seven-membered N-heterocyclic carbene (7NHC) ligands for palladium is described. These 7NHC ligands are derived from enatiomerically pure 2,2′-diamino-6,6′-dimethylbiphenyl, which is transformed via a synthetic sequence consisting of ortho-arylation, N-alkylation, and cyclization to afford seven-membered-ring amidinium salts. Synthesis of the 7-membered amidinium salts benefits from microwave irradiation, and in-situ metalation of the amidinium salts yields 7NHC-PdII complexes. The chiral 7NHC-Pd complexes were examined as chiral catalysts under aerobic conditions in two intramolecular oxidative amination reactions of alkenes. In one case, enantioselectivities up to 63% e.e. were obtained, while the other substrate underwent cyclization to afford essentially racemic products. The catalytic data compare favorably to results obtained with a PdII catalyst bearing a chiral five-membered-ring NHC ligand and, thereby, highlight the potential significance of this new class of chiral NHC ligands.
Chiral monodentate N-heterocyclic carbenes (NHCs) have broad potential utility in asymmetric catalysis, but successful applications of such ligands remain somewhat limited.1,2 To our knowledge, only six transformations have been reported in which chiral monodentate NHC ligands have been employed to afford enantioselectivities above 90% e.e. (for representative examples, see Figure 1). The transformations include olefin metathesis, 3 1,4-addition of arylboronic acids to enones, 4 hydrosilylation of ketones, 5 intramolecular alpha-arylation of amides,6,7 copper-catalyzed conjugate addition to cycloheptenone with diethylzinc,8 and Ni-catalyzed reductive coupling of 1,2-dienes and aldehydes with triethylsilane9. All of these transformations utilize 5-membered-ring NHC ligands. Two strategies have been employed in the design of chiral monodentate 5-membered NHCs: “chiral relay” and the use of chiral nitrogen substituents. In the first strategy, developed by Grubbs and coworkers,3 chirality in the N-heterocycle induces an asymmetric conformation of the nonsymmetrical (but achiral) nitrogen substituents (Figure 1, ligand L1). In the second strategy, the ligand chirality arises from the incorporation of chiral nitrogen substituents into the ligand, for example, by Buchwald-Hartwig amination of chiral cyclophanes (Figure 1, ligand L2)5,4 or by reductive amination of 1,2-diones with chiral amines (Figure 1, ligands L3–L5).6,7,8 Although attempts have been made to develop ligands that feature both chiral relay and chiral nitrogen substituents, only modest success has been achieved thus far.8,10
NHCs are excellent ligands for metal-catalyzed aerobic oxidation reactions because metal-NHC complexes are oxidatively stable.11 NHC-Pd complexes have been employed as catalysts for aerobic oxidation of alcohols, 12 Wacker-type cyclization of ortho-allyl phenols, 13 and intramolecular oxidative amination of olefins (Figure 2a, b, and c, respectively).14 We have been interested in the development of enantioselective aerobic oxidative amination reactions of alkenes, and the use of chiral NHC ligands was a logical starting point. In connection with these efforts, we have been exploring the development and application of a new class of NHCs based upon a seven-membered heterocyclic framework (Figure 3b).15,16 These 7-membered NHCs (7NHCs) are attractive because they feature an axially-chiral heterocyclic framework that provides an out-of-plane projection of the nitrogen substituents (Figure 3b). This non-planarity contrasts the in-plane orientation of the nitrogen substituents in 5-membered NHCs (Figure 3a), and we reasoned that increased enantioselectivity might be achieved in asymmetric catalysis with 7NHCs. We have reported that a racemic 7NHC-Pd complex is an active catalyst for the intramolecular aerobic oxidative amination of o-crotyl tosylanilide (Figure 2c),14 but, at the time of this work, we had not yet succeeded in preparing enantiomerically resolved 7NHC-Pd complexes. Our initial synthesis of 7NHC-Pd complexes15 is not amenable to the synthesis of resolved analogs. Here, we describe the synthesis of enantiomerically resolved seven-membered-ring amidinium salts, metalation of the amidinium salts by PdII and the investigation of 7NHC– PdII complexes in intramolecular aerobic oxidative heterocyclization reactions. The enantioselectivity data from the catalytic reactions is modest (≤ 63% e.e.). Nevertheless, the 7NHC ligand proves to be substantially better than one of the most successful known 5NHC ligands.
At the time this work began, there were only two reported examples of chiral monodentate 5NHCs employed in reactions that achieved e.e.'s above 90%.3c,4 The chiral 5NHCs developed by Grubbs et al. (L1, Figure 1)3 were not useful for our goal of accessing NHC-Pd(O2CR)2 catalysts for asymmetric aerobic Wacker-type oxidation reactions because such catalysts were not synthetically accessible.17 The chiral 5NHCs developed by Andrus et al. (L2, Figure 1)4 proved to be prohibitively difficult to synthesize. Within this context, we turned our attention to the development of new chiral NHC ligands, and our efforts led to the 7NHC derivatives. After achieving preliminary success in the preparation of racemic 7NHC-Pd complexes,15 we envisioned two potential strategies to access resolved derivatives of 7NHC ligands: (1) incorporation of substituents ortho to the nitrogens (the 3 and 3′ positions of the biphenyl moiety), and (2) use of resolved biphenyl diamines (i.e. biphenyl diamines with substituents in the 6 and 6′ positions). In the first strategy, we reasoned (and obtained preliminary support from computational modeling) that steric interactions between the ortho groups and the nitrogen substituents should prevent racemization of the 7NHC. This strategy proved unattractive, however, because it would require independent resolution of each new amidinium salt (the 7NHC precursor).18 We therefore focused our efforts on executing the second strategy, which would allow many resolved 7NHC complexes to be synthesized from a single stock of resolved diamine.
As we have reported previously,15a the use of 2,2′-diamino-1,1′-binaphthyl (DABN) for the synthesis of resolved DABN-derived seven-membered-ring amidinium salts was successful only with neopentyl groups as nitrogen substituents (e.g. Figure 4a); the corresponding amidinium salt bearing N-(2-adamantyl) or N-aryl groups were not accessible (e.g. Figure 4b).15, 19 These observations were problematic because palladium complexes bearing 7NHCs with primary nitrogen substituents proved to be unstable. The instability of 7NHC-Pd complexes bearing primary alkyl nitrogen substituents appears to arise from decreased steric protection of the carbene carbon relative to those with secondary nitrogen substituents. Based on this hypothesis, we reasoned that substituents in the ortho positions (i.e., the 3 and 3′ positions of the biphenyl moiety) could provide additional steric protection of the carbene carbon and, thereby, permit primary alkyl nitrogen substituents to be employed (Figure 4c). Based on this proposal, we explored methods to access ortho substituted biaryl diamines.
Installation of steric bulk ortho to the nitrogen substituents of biaryl diamines required an ortho functionalization strategy. We recently reported that application of an ortho C–H arylation reaction developed by Daugulis and Zaitsev20 provides ready access to ortho diarylated biaryl diamines (Scheme 1).21 An important result from this study is that the ortho-arylation reaction is more effective for 2,2′-diamino-6,6′-dimethylbiphenyl (4) than for DABN, and we therefore turned our attention completely to the resolved diamine 4.22 We next examined the synthesis of [7NHC-H]+ salts derived from these new resolved quaterphenyl diamines.
This synthetic access to ortho-functionalized biaryldiamines enabled us to test the hypothesis that primary nitrogen substituents could be used in the preparation of stable NHC-Pd complexes. Our initial studies were carried out on the inexpensive and readily available diamine 6, which was synthesized by acetylation, Dagulis-Zaitsev arylation, and deacetylation of 2,2′-diaminobiphenyl.21 Sequential addition of 1-naphthaldehyde and lithium aluminum hydride to the quaterphenyl diamine 621 furnished the di-N-alkylated quaterphenyl diamine 7 in high yield. Cyclization of this diamine afforded the desired amidinium tetrafluoroborate salt 8. Addition of [Pd(allyl)Cl]2 and KOtBu to 8 furnished the stable NHC-Pd(allyl)Cl complex 9, albeit in low yield (Scheme 2). The structure of 9 was verified by X-ray crystallographic analysis (Figure 5).23 Synthesis of the stable NHC-Pd complex 9 bearing primary nitrogen substituents supported our hypothesis that incorporation of steric bulk in the ortho-positions of the biphenyl diamine could provide steric protection of the NHC and permit the synthesis of stable NHC-metal complexes bearing primary nitrogen substituents.
The next important challenge was to demonstrate that this synthetic approach could be used with resolved biphenyl diamines. Toward this end, (S)-5 was alkylated via sequential reductive amination with cyclohexane carboxyladehyde and LiAlH4 to afford (S)-10 in 94% yield (Scheme 3). Initial attempts to cyclize (S)-10 using published conditions15,24 with HC(OEt)3 and NH4BF4 to form amidinium salt (S)-11 were largely unsuccessful, providing a thick dark reaction mixture after prolonged reaction times, from which only a small amount of nearly pure amidinium salt could be extracted; shorter reaction times provided very low yields as well. We ultimately found the cyclization reaction could be achieved by employing microwave irradiation.25 The amidinium salt (S)-11 was obtained in 38% yield, and workup of the crude mixture was not as challenging. Metalation of amidinium salt (S)-11 with KOtBu and [Pd(allyl)Cl]2 afforded the 7NHC-Pd(allyl)Cl complex (S)-12 in 29% yield (Scheme 3). As NHC-Pd(carboxylate)2 complexes are of catalytic relevance,11,12 we sought to make such a complex from (S)-12. Addition of ethereal HCl to (S)-12 to afford the [NHC-PdCl2]2 complex (S,S)-13 followed by addition of AgO2CCF3 (AgTFA) afforded the NHC-Pd(TFA)2(OH2) complex (S)-14 (Scheme 4). The molecular structure of (S)-14 was confirmed in the solid state by X-ray diffraction analysis (Figure 6).23 Unfortunately, the yields of amidinium salt and NHC-Pd(allyl)Cl are quite low, despite extensive attempts to improve them. 26 Nonetheless, these results represented the first successful preparation of resolved 7NHC complexes of this type.
The [NHC-PdCl2]2 complex (S,S)-13 was first examined as a catalyst for the asymmetric aerobic oxidative cyclization of substrate 15 (Table 1).27 This substrate was examined because it is the only substrate reported to undergo palladium-catalyzed asymmetric aerobic oxidative amination.28 Yang and coworkers achieved up to 86% e.e. in the oxidative cyclization of this substrate with a (–)-sparteine-Pd catalyst. NHC ligands have not been tested previously in this oxidative cyclization reaction. Representative screening data in Table 1 reveal only modest success with the use of the 7NHC-Pd complex (S,S)-13 as the catalyst. In the best case, the product 16 was obtained in 63% e.e., albeit in rather low yield (Table 1 entry 3). Other conditions examined resulted in significantly lower e.e. or no formation of 16. For example, base was shown to facilitate substrate oxidation by PdII, but iPr2NEt was the only base among those tested (including Na2CO3, NaOAc, NaO2CPh, MgO, Ca(OH)2, NEt3) that led to even modest levels of catalytic turnover. The origin of these observations is not currently known.
We next turned our attention to the aerobic oxidative amination of substrate 2, which has been shown by us to be an effective substrate for aerobic oxidative amination catalyzed by NHC-Pd complexes, including one of our racemic 7NHC ligands (Figure 2).14 The conditions employed in entries 1-6 of Table 2 resemble the best conditions for the reported racemic cyclization of 2,14 whereas the conditions employed in entry 7 mirrored the best conditions for the cyclization of 15. Although modest yields of product 3 were obtained, the cyclization product was nearly racemic in every case (Table 2).
When the oxidative cyclization studies described above were nearly complete, Kündig and coworkers reported the use of chiral 5NHC ligands in Pd-catalyzed intramolecular α-arylation reactions (Figure 1d).6,7 The latter work, which provided the first highly-enantioselective reaction (>90% e.e.) employing a chiral NHC-Pd catalyst, prompted us to examine the most successful 5NHC ligand from this work in Wacker-type oxidative cyclization reactions.
Synthesis of palladium complexes derived from the amidinium iodide salt 1 (Figure 1d) were carried out by combining 1 with [Pd(allyl)Cl]2 and KOtBu to afford the NHC-Pd(allyl)Cl complex 17, followed by protonolysis of the allyl ligand with ethereal HCl to afford the [NHC-PdCl2]2 complex 18.29 Anion metathesis with silver(I)-carboxylate salts did not furnish the desired NHC-Pd(carboxylate)2 complexes, but instead led to formation of the cyclometalated complexes 19 and 20 (Scheme 5). The solid-state molecular structure of 19 was unambiguously determined by X-ray diffraction analysis (Figure 7).23 Although complexes 19 and 20 were not the targeted NHC-Pd(carboxylate)2 catalysts, we nonetheless tested these complexes as catalysts for the oxidative cyclization reactions of substrates 2 and 15.
In the aerobic oxidative cyclization of substrate 15 with catalysts 19 and 20, improved yields could be obtained relative to those with catalyst (S,S)-13; however, the product 16 was racemic in every case (Table 3). The aerobic oxidative cyclization of 2 with catalysts 19 and 20 proceeded in excellent yield (Table 4), but, once again, the product was always racemic. The improved yields with the 5NHC-Pd catalysts may reflect the smaller size of the 5NHC ligand compared to the 7NHC ligand.
In this study, we have achieved preparation of the first enantiomerically resolved 7NHC-Pd complexes and investigated their reactivity in aerobic oxidative cyclization reactions of alkenes. The 7NHC-precursor amidinium salt is accessible through a sequence involving a Pd-catalyzed ortho-arylation of the biaryl ring, incorporation of nitrogen substituents via reductive amination, and cyclization using HC(OEt)3 and NH4BF4 under microwave irradiation. The ortho-arylation reaction is critical to access stable 7NHC-Pd complexes. These ligands were employed with mixed success in Pd-catalyzed aerobic oxidative cyclization reactions. In one case, up to 63% e.e. was achieved in a Pd-catalyzed intramolecular oxidative amination reaction. Although this level of enantioselectivity is relatively modest, the result is significantly better than that obtained with Pd-catalysts bearing chiral five-membered NHC ligands (≤7% e.e.). These observations highlight the prospects for the use of chiral seven-membered NHCs of the type described here in asymmetric catalysis.
Reactions which were performed under nitrogen were run in column-dried solvents,30 whereas reactions performed in a glove box were run in ketyl-dried solvents. 1H and 13C NMR spectra were recorded on 300 or 500 MHz spectrometers. 1H chemical shifts (δ) are reported in ppm relative to SiMe4 (0.0 ppm) while 13C chemical shifts are reported in ppm relative to deuterated solvent (77.23 ppm for CDCl3 or 128.06 for benzene-d6).
Screening was conducted in disposable Borosilicate Heavy Wall 13 × 100 mm Culture Tubes (Fisherbrand). Catalyst and solid additives were weighed into reaction tubes, which were then loaded onto a custom 48-well parallel reactor mounted on a Large Capacity Mixer (Glas-Col). The headspace was purged under a positive flow of dioxygen with mild vortexing for fifteen minutes. A stock solution of the substrate in toluene (0.1M) was added to the reaction tubes, which were then vortexed for 18 h under 1 atm dioxygen at 80 °C, after which time the reactions were filtered over celite and the solvent was removed from each tube under reduced pressure prior to analysis by 1H NMR spectroscopy, gas chromatography, and/or HPLC.
470 mg diamine 6 was combined with 778 μL 1-naphthaldehyde and a few crystals of p-TsOH in ~250 mL toluene. A Dean-Stark apparatus was attached, and the reaction was heated to reflux for 48 h. Solvent was then removed in vacuo and the residue was taken up in 25 mL dry THF. 339 mg LiAlH4 (6.4 equiv) was added slowly and the solution was heated to 50 °C for two hours. After cooling to room temperature, the reaction was carefully quenched by sequential addition of 0.3 mL H2O, 0.6 mL 10% NaOH(aq), then 0.9 mL H2O. The solid that formed was removed by filtration, and the filtrate was concentrated in vacuo. Product was purified by column chromatography (SiO2 packed with 5% NEt3 in toluene, then flushed with toluene before loading crude) using a gradient eluent (pure toluene to 1:1 toluene-EtOAc) to afford 605 mg 7 as a light yellow oil (70% yield). HRMS: m/z (ESI) calculated [MH]+ = 617.2956, measured 617.2963 (Δ = 1.1 ppm). 1H NMR (300 MHz, CDCl3) 3.99 (s, 4H) 6.90 (d, J = 6.0 Hz, 2H), 7.01 (t, 7.5 Hz, 2H), 7.14 - 7.36 (m, 22H), 7.60 (d, J = 8.4 Hz, 2H), 7.69 (d, J = 9.0 Hz, 2H). 13C NMR (75MHz, CDCl3) 50.6, 121.8, 123.6, 125.4, 125.6, 126.3, 127.2, 127.9, δ128.5, 128.5, 129.4, 130.9, 131.1, 132.5, 133.8, 134.2, 135.6, 140.9, 145.0.
3.1 g diamine 7 was combined with 560 mg NH4BF4 in 150 mL HC(OEt)3. The reaction was heated to 130 °C for 48 h. The reaction was then concentrated in vacuo to a thick oil and taken up in minimal methylene chloride. Excess hexanes were added to drive the product out of solution, and the suspension was cooled in a freezer overnight. Tan solid was collected by filtration and washed with pentane. The solid was then taken up in minimal refluxing ethanol and cooled in a freezer. The resultant powder was collected by filtration, washed with pentane, and dried under high vacuum to afford 1.7 g pure 8 as a tan solid (47% yield). HRMS: m/z (ESI) calculated [MH]+ = 627.2795, measured 627.2805 (Δ = 1.6 ppm). 1H NMR (300 MHz, CDCl3) 4.35 (d, J = 14.4 Hz, 2H), 4.82 (d, J = 14.4 Hz, 2H), 6.73 (br. s, 4H), 7.04 (ddd, J = 8.1, 6.9, 1.2 Hz, 2H), 7.10 - 7.15 (m, 4H), 7.21 (ddd, J = 8.4, 6.9,1.2 Hz, 2H), 7.28 - 7.31 (m, 4H), 7.36 - 7.54 (m, 10H), 7.61 (d, J = 7.8 Hz, 2H), 7.68 (d, J = 8.4 Hz), 9.49 (s, 1H). 13C NMR (75MHz, CDCl3) 57.1, 122.2, 125.9, 126.0, 126.7, 128.0, 128.9, 129.1, 129.3, 129.6, 130.1, 130.2, 132.2, 132.4, 133.8, 134.3, 136.9, 137.2, 142.1, 179.1.
756 mg 8 (1.06 mmol), KOtBu (132.2 mg, 1.1 equiv), and [Pd(allyl)Cl]2 (230 mg, 0.60 equiv), were combined in 30 mL dry THF in a glove box. The reaction was stirred at room temperature for 20 h, after which it was removed from the glove box. The suspension was filtered over celite and the solvent removed in vacuo. 9 was purified by column chromatography (SiO2) using a gradient eluent (40% Et2O in hexanes to pure Et2O) as a lightly yellow solid (100 mg, 12% yield). The product is a 1.5:1 mixture of allyl rotamers by 1H NMR spectroscopy. HRMS: m/z (ESI) calculated [M(-Cl)]+ = 771.2169, measured 771.2178 (Δ = 1.2 ppm). 1H NMR (300 MHz, CDCl3) (all peaks are quite broad. Identity is known from X-ray crystal structure, but NMR spectra are difficult to interpret) 1.64 (m, 1.5H), 2.25 (m, 1H), 2.38 (m, 1H), 2.84 (m, 1.5H), 3.31 (m, 2.5H), 4.29 (m, 1.5H), 4.72 (m, 1H), 5.00 (m, 1.5H), 5.24 (m, 1H), 6.94 - 7.64 (m, 45H). 13C NMR (75MHz, CDCl3) 58.2 (br. peak), 115.7, 116.4, 125.0, 125.7, 127.7, 127.9, 128.4, 129.1, 132.0, 133.4, 136.2, 140.1, 144.9 (br. peak).
2.0 g diamine (S)-5 was combined with 1.7 mL cyclohexane carboxaldehyde (2.5 equiv) and a few crystals of p-TsOH in 120 mL toluene in a Dean-Stark apparatus. The reaction was heated to reflux for 60 h, cooled to room temperature, and solvent was removed in vacuo. The resultant residue was taken up in 50 mL dry THF and 650 mg LiAlH4 was carefully added portionwise. The reaction was stirred at 50 °C for 2 h, cooled to room temperature, and carefully quenched by sequential addition of 1 mL H2O, 2 mL 10% NaOH(aq), and 3 mL H2O. The solid which had crashed out was removed by filtration, and solvent was removed from the filtrate in vacuo to afford (S)-10 in 95% yield. HRMS: m/z (ESI) calculated [MH]+ = 557.3891, measured 557.3883 (Δ = 1 ppm). 1H NMR (300 MHz, CDCl3) 0.41 (br. q, J = 12.3 Hz, 4H), 0.90 - 1.09 (m, 8H), 1.20 - 1.28 (m, 4H), 1.48 - 1.55 (m, 4H), 2.02 (s, 6H), δ2.27 (m, 4H), 3.27 (br. s, 2H), 6.87 (dd, J = 7.8, 0.8 Hz, 2H), 7.10 (d, J = 7.6 Hz, 2H), 7.27 - 7.30 (m, 2H), 7.35 - 7.40 (m, 4H), 7.49 - 7.52 (m, 4H). 13C NMR (75MHz, CDCl3) 20.5, 26.1, 26.7, 30.9, 31.0, 38.8, 54.5, 121.9, 126.5, 127.6, 128.4, 129.1, 131.3, 137.1, 142.4, 145.8.
Diamine (S)-10, NH4BF4 (1.3 equiv), and 100 mL HC(OEt)3 were combined and subjected to microwave irradiation (equilibration time to 100 °C = 10 minutes, 2 h at 100 °C, max power = 250 watts). Concentration on a rotary evaporator, dissolving in minimal hot toluene, and crashing out with hexanes afforded some powder and a sticky orange semi-solid. The powder was filtered away, and the semi-solid was again taken up in hot toluene and crashed out with hexanes. This procedure was repeated as necessary until no more powder formed. Dissolved filtered powder in dichloromethane and filtered to remove excess NH4BF4. Removed solvent from filtrate to afford a sticky solid, which was re-dissolved in minimal dichloromethane and added dropwise to a 250 mL graduated cylinder filled with hexanes (powder forms immediately on addition of each drop). After complete addition, let suspension sit for 30 minutes then decanted the hexanes. Dissolved solid in dichloromethane and transferred to a flask. Solvent was removed in vacuo to afford 38% yield (1.35 g) of (S)-11 as a tan powder. HRMS: m/z (ESI) calculated [MH]+ = 567.3734, measured 567.3736 (Δ = 0.4 ppm). 1H NMR (300 MHz, CDCl3) 0.39 (qd, J = 11.7 Hz, 2H), 0.52 - 0.71 (m, 4H), 0.81 - 0.97 (m, 4H), 1.10 - 1.35 (m, 6H), 1.39 -1.43 (m, 6H), 2.29 (s, 6H), 2.93 (dd, J = 13.8, 5.1 Hz, 2H), 3.37 (14.0, 6.1 Hz, 2H), 7.27 - 7.35 (m, 6H), 742 - 7.49 (m, 4H), 7.55 - 7.61 (m, 4H), 9.05 (s, 1H). 13C NMR (75MHz, CDCl3) 19.6, 25.6, 25.7, 25.8, 29.7, 30.0, 40.1, 61.4, 127.7, 128.9, 130.1, 131.0, 132.6, 132.8, 134.4, 137.6, 138.3, 145.4, 175.8.
This complex was synthesized analogously to compound 9. Product was purified by column chromatography (SiO2) using 1:1 Et2O-hexanes as the eluent. Product was isolated in 29 % yield (195 mg) as a lightly yellow solid. Peaks in NMR spectra are challenging to assign, and this structure is assigned from previous precedent, 1H NMR spectroscopy (d.r. ~ 2.2 : 1), and from reactivity ((S)-14 is derived quantitatively from this species, which has been fully characterized using ESI-MS, 1H and 13C NMR spectroscopy, and single-crystal X-ray diffraction analysis). 1H NMR (300 MHz, CDCl3) 0.47 - 1.40 (m, 31.9H), 2.15 - 2.53 (m, 10.15H), 2.53 - 3.05 (m, 4.8H), 3.34 - 3.65 (m, 4.25 H), 4.33 (m, 2H), 5.23 (m, 0.45H), 5.39 (m, 1H), 7.09 - 7.37 (m, 20.3H).
195 mg (S)-12 (0.26 mmol) was dissolved in 50 mL Et2O, and 3.25 mL 2.0 M ethereal HCl (25 equiv) was added via syringe. This mixture was stirred at room temperature for 6 h, over which time the solution turned bright yellow. Removed solvent in vacuo and dried under high vacuum to afford 195 mg of a bright yellow compound which reacts as the drawn compound when combined with AgO2CCF3 (below). 1H NMR spectroscopy revealed two large broad peaks, in the alkyl and aryl regions, but no distinct peaks were observed. Peaks in NMR spectra are challenging to assign, and this structure is assigned from previous precedent,15 and from reactivity ((S)-14 is derived quantitatively from this species, which has been fully characterized using ESI-MS, 1H and 13C NMR spectroscopy, and single-crystal X-ray diffraction analysis).
83 mg (S,S)-13 (0.056 mmol) was combined with 54.2 mg (4.4 equiv) AgO2CCF3 in dichloromethane. This reaction mixture was stirred at room temperature for 1.5 h, during which time a white solid was observed to have formed. Filtration over celite and removal of solvent under vacuum afforded (S)-14 as a tan-yellow solid in quantitative yield. The identity of this species was confirmed by X-ray crystallographic analysis (Figure 11). ESI-MS: m/z (major peak containing Pd) 707.7 ([NHC-Pd(OH2)(OH)]+). 1H NMR (300 MHz, CDCl3) 0.39 - 0.51 (m, 4H), 0.83 - 1.44 (m, 18H), 2.19 (s, 6H), 3.23 (dd, J = 13.5, 5.1 Hz, 2H), 4.32 (dd, J =13.2, 8.1 Hz, 2H), 7.18 (d, J = 7.5 Hz, 2H), 7.31 (d, J= 7.5 Hz, 2H), 7.46 - 7.52 (m, 10H). 13C NMR (75MHz, CDCl3) 19.6, 25.5, 25.9, 26.1, 29.8, 30.8, 38.4, 60.3, 128.5, 128.6, 129.4, 129.7, 132.0, 133.3, 134.3, 137.6, 139.4, 145.1, 201.8. [α]D25 +3.6 (c 5.5, CH2Cl2).
In a glove box, amidinium salt 1 (285 mg, 0.552 mmol), [Pd(allyl)Cl]2 (131 mg, 0.65 equiv), and KOtBu (80 mg, 1.3 equiv) were combined in 20 mL THF. The reaction stirred at room temperature in the dark for 12 h, after which time the reaction flask was removed from the glove box and exposed to ambient conditions. Solvent was removed on a rotary evaporator, and product was isolated by column chromatography (SiO2) using 1 : 1 Et2O-hexanes as the eluent to afford a light tan solid with a 1H NMR spectrum which could be consistent with the drawn product, but was not conclusive. Carried material forward without full characterization. Identity is based on formation of 19 and 20.
To crude 17 was added 1 mL ethereal HCl (2.0 M, 3.6 equiv compared to amidinium salt 1). The reaction was stirred at room temperature for 2h. Solvent was removed on a rotary evaporator to afford 141 mg 18 as an orange-brown solid in 45% yield over two steps. Again, the 1H NMR spectra was consistent with the drawn product, but inconclusive. Assignment is based on formation of 19 and 20. 1H NMR (300 MHz, CDCl3) (all peaks are broad) 1.06 - 1.16 (m, 36H), 3.29 (s, 6H), 3.34 (s, 6H), 6.92 (s, 2H), 7.07 - 7.23 (m, 12H), 7.38 (s, 2H), 7.66 (s, 4H).
81 mg 18 (0.072 mmol) was combined with 49 mg AgOAc (4.1 equiv) in dichloromethane. The reaction stirred at room temperature in the dark for two hours, after which time the reaction was filtered through celite to remove the AgCl. Solvent was removed in vacuo to afford 74 mg 6 (86% yield) as a brown solid. The structure of 19 was unambiguously assigned by X-ray diffraction analysis (Figure 4): X-ray-quality crystals were obtained from slow evaporation of a CDCl3-solution of 6 in a round-bottom flask. An X-ray-quality crystal of 6 was re-dissolved in CDCl3, and gave a 1H NMR spectrum which was identical to the bulk material. HRMS: m/z (ESI) calculated [M-OAc]+ = 491.1836, measured 491.1830 (Δ = 1.0 ppm). 1H NMR (300 MHz, CDCl3) 1.09 (s, 9H), 1.26 (s, 9H), 2.11 (br. s, 6H), 2.32 (s, 3H), 2.94 (s, 3H), 5.13 (s, 1H), 6.26 (s, 1H), 6.79 - 6.99 (m, 2H), 6.99 (d, J = 2.1 Hz, 1H), 7.10 - 7.16 (m, 4H), δ7.37 - 7.48 (m, 2H), 7.49 (dd, J = 5.4, 4.1 Hz, 1H). 13C NMR (75MHz, CDCl3) 21.8, 22.1, 27.4, 29.3, 37.3, 37.7, 66.1, 71.8, 118.0, 123.0, 125.5, 126.1, 127.2, 127.8, 128.2, 131.4, 131.5, 134.0, 135.7, 136.4, 138.9, 139.2, 162.2, 174.2.
Prepared from 18 analogously to the above synthesis of 19, using AgO2CCF3 in place of AgOAc. Product was obtained in ~100% yield (67 mg) as a dark brown solid. HRMS: m/z (ESI) calculated [M]+ = 491.3426, measured 491.3437 (Δ = 2.2 ppm). 1H NMR (300 MHz, CDCl3) (all peaks are rather broad) 1.03 - 1.29 (m, 36H), 2.33 (m, 6H), 2.51 (m, 6H), 5.84 - 5.92 (m, 4H), 7.03 - 7.91 (m, 16H), 8.89 (br. s, 2H). 13C NMR (75MHz, CDCl3) 15.5, 21.4, 21.7, 27.1, 27.4, 27.6, 28.2, 38.3, 39.4, 66.1, 66.9, 72.4, 118.7, 123.1, 124.7, 125.0, 127.3, 128.4, 129.3, 129.6, 132.0, 133.2, 134.1, 137.0, 137.3, 146.7, 147.4.
CCS is grateful to the American Chemical Society Division of Organic Chemistry for a graduate fellowship. This work was funded by the National Institutes of Health (R01 GM67173) and a UW-Madison Graduate School Technology Transfer Grant.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.