Chiral 1,2-diamino compounds are important building blocks in organic chemistry for biological applications and as asymmetric inducers in stereoselective synthesis that are challenging to prepare in a straightforward and stereoselective manner. Herein, we disclose a cost-effective and readily available Cu-catalyzed system for the reductive coupling of a chiral allenamide with N-alkyl substituted aldimines to access chiral 1,2-diamino synthons as single stereoisomers in high yields. The method shows broad reaction scope and high diastereoselectivity and can be easily scaled using standard Schlenk techniques. Mechanistic investigations by density functional theory calculations identified the mechanism and origin of stereoselectivity. In particular, the addition to the imine was shown to be reversible, which has implications toward development of catalyst-controlled stereoselective variants of the identified reductive coupling of imines and allenamides.
Chiral 1,2-diaminocompounds are important building blocks in organicchemistry for biological applications and as asymmetric inducers in stereoselective synthesis that are challenging to prepare in a straightforward and stereoselective manner. Herein, we disclose a cost-effective and readily available Cu-catalyzed system for the reductive coupling of a chiral allenamide with N-alkyl substituted aldimines to access chiral 1,2-diamino synthons as single stereoisomers in high yields. The method shows broad reaction scope and high diastereoselectivity and can be easily scaled using standard Schlenk techniques. Mechanistic investigations by density functional theory calculations identified the mechanism and origin of stereoselectivity. In particular, the addition to the imine was shown to be reversible, which has implications toward development of catalyst-controlled stereoselective variants of the identified reductive coupling of imines and allenamides.
Chiral vicinal diamines
are extremely valuable and important motifs
in organicchemistry that are exploited by both nature and the pharmaceutical
industry for their biological activities,[1,2] and
in stereoselective organic synthesis as powerful chiral inducers through
application as organocatalysts,[3] chiral
ligands[4] for transition metalcatalyzed
reactions, and as chiral auxiliaries.[5] For
example, a variety of biologically active pharmaceuticals and natural
products are given in Figure possessing either the chiral 1,2-diamino-fragment or its
corresponding urea form.[2] Representative
therapeutics being developed for the treatment of important human
diseases include antibiotics (penicillin,[2e] jogyamycin[2j]), anticancer compounds (cisplatin
derivatives,[6] LP99[2c]), HIV protease inhibitors (NBD-11021[2d]), NK1-antagonists[7] (CP-99,994;[2a] Sch425078[2g]) for
central-nervous-system (CNS) related diseases and rheumatoid arthritis,
and influenza (tamiflu).[2b]
Figure 1
Selected examples of
chiral 1,2-diamine- and urea-derived biologically
active molecules.
Selected examples of
chiral 1,2-diamine- and urea-derived biologically
active molecules.Due to the biological
and synthetic value of chiral 1,2-diamines,
stereoselective methods for their preparation are an important endeavor
in organicchemistry.[1a,1c,1d,8] Potential synthetic options to access the
chiral vicinal diamine moiety can be envisioned to occur either by
formation of the two C–N bonds starting from unsaturated hydrocarbons
(2, Scheme A) or through direct C–C bond formation between C1 and C2
of the 1,2-diamine from two N-substituted reagents
(Scheme B).[1a,1c,1d,8] Using
a C–N bond forming approach (Scheme A), diamination may be achieved by forming
both C–N bonds at the same time,[8,9] or sequentially
through either aziridination[10] followed
by ring-opening with an amine nucleophile[1a,1c,1d,11] or through
aminohydroxylation[12] followed by alcohol
activation and amine substitution.[1a−1d] While direct catalytic 1,2-diamination
of 2 represents an ideal strategy for diamine synthesis,
the amino-groups added across the π-system are typically identical
leading to the formation of diamines with identical substituents (i.e.,
R3 = R4 in 1),[8,9] and
a recent approach employing electrochemistry[9b] suffers from potentially forming high-energy/explosive diazocompounds[13] en route to the desired diamines. Additionally,
the aziridination/ring-opening strategy can suffer from poor stereoselectivity
in the aziridination step and regiochemistry issues in the subsequent
opening step, while the aminohydroxylation route requires regiocontrol
in the aminohydroxylation step followed by additional transformations
to convert 4 to the desired diamine. Alternatively, synthesis
of 1 through C–C bond formation can be achieved
through aza-pinacolcoupling of two imines,[14] nitro-Mannich,[15] or glycine-Mannich[16] reactions (Scheme B). Typical aza-pinacolcoupling protocols
only afford symmetrical diamines through homocoupling of a single
imine; however, recent photoredox strategies[14h−14l] enabling the generation of α-aminoradicals[17] from amines have enabled cross-selective coupling of imines
and N-methylamines.[14j−14l] Furthermore, nucleophilic
additions to imines using α-aminoanion derivatives[18] from nitroalkanes (7)[15] or protected glycines (8)[16] offer another entry into the diaminecore 1.
Scheme 1
Synthetic Strategies toward the Synthesis of 1,2-Diamines
In regards to chiral amine synthesis, asymmetric
allylation of
imines using allyl organometallic nucleophiles (10) by
direct addition or through catalyst control has been an area of intense
research in organicchemistry (Scheme ).[19] The chiral allylamine
products (11) are highly valuable in the context of the
synthesis of complex amine-containing organiccompounds because of
the high versatility of the olefin functional group present within 11. Substituted allylorganometallic reagents (e.g., 12) allow for increased molecular complexity by introducing
two stereocenters in the allyl addition reaction (e.g., 13, Scheme B). Therefore,
we envisioned that use of an amino-substituted allyl reagent 12 in addition reactions with imine electrophiles would be
a powerful strategy to prepare 1,2-diamines (13) with
differential substitution patterns on nitrogen and containing an olefin
motif for further functional group manipulations. Surprisingly, only
a single example of such a strategy for the preparation of 1,2-diamines
has been reported, which employs a lithiated derivative of 12 (M = Li) with chiral tert-butanesulfinimide
derived aldimines affording products in moderate yields with mixtures
of branched and linear allylation products.[20] In contrast, amino-substituted allyl reagents 12 have
been used in reactions employing carbonyl electrophiles to provide1,2-aminoalcohols (16).[21−23] Recently, the Krische[22] group and our own lab[23] have developed reductive coupling[24,25] procedures
for the catalytic generation of amino-substituted allyl reagents 12 and have studied their reactions with carbonyl electrophiles
(Scheme C). These
techniques represent orthogonal methodologies whereby the Krische[22a] system employs a chiral Ir-catalyst and processes
aldehyde electrophiles using an achiral allenamide (15), while our work utilizes a Cu-catalyst and a chiral allenamide
(15a) for reactions using ketone electrophiles.[23] Based on our success in the stereoselective
Cu-catalyzed reductive coupling of ketones and chiral allenamides
to afford branched chiral 1,2-aminoalcohols 16(23a) or the corresponding linear products,[23b] and the lack of literature data for imine allylation
reaction utilizing amino-substituted allylic nucleophiles, we began
to investigate the reaction of allenamide 15a with imine
electrophiles 9 for the stereoselective synthesis of
chiral 1,2-diaminesynthons 17 (Scheme D). The results of these studies leading
to the identification of a practical and highly stereoselective synthesis
of diaminesynthons 17 using Cu-catalyzed reductive coupling
are disclosed herein.
Scheme 2
Proposed Allylation Strategy toward the
Synthesis of 1,2-Diamines
Results
and Discussion
Reaction Optimization
To investigate
the proposed Cu-catalyzed
reductive coupling of imines and allenamides, initial studies examined
the ligand effect when employing DMB-protected imine 9a with chiral allenamide 15a in the reaction (Table ). The phenyl-derived
Evans oxazolidinone of allenamide 15a was specifically
targeted due to its low-cost and high-availability,[26] and because it allows for more deprotection options of
the desired diamine products over other alkyl-substituted oxazolidinones
(i.e., hydrogenolysis). The DMB-group of the aldimine was employed
due to its acid lability to allow for chemoselective differentiation
of the two amine protecting groups in the final products (18a/19a). Gratifyingly, a variety of phosphine ligands
(entries 1–7) afforded urea product 19a presumably
resulting from migration of the carbamatecarbonyl (Scheme ), whereas an N-heterocyclic carbene (NHC) ligand provided poor conversion (entry
8). In all cases, a single diastereomer of product was obtained as
determined by 1HNMR spectroscopy of the unpurified reaction
mixture. Notably, the bidentate phosphinedcpe that has been utilized
previously in Cu-catalyzed reductive coupling of C-substituted allenes and imines[25c] afforded
only a moderate yield with a substantial amount of unreacted imine
(20%, entry 1). Monodentate phosphine ligands (entries 2–7)
worked well with the exception of sterically demanding ligands that
afforded poor conversion of the imine (entries 3, 4). Ultimately,
the use of PCy3 as ligand afforded the highest yield of 19a in the reaction (entry 2). Use of solvents other than
toluene in the reaction (entries 9–12) offered no improvements.
Finally, addition of 2 equiv of t-BuOH to the reaction
led to the exclusive formation of 18a in excellent yield
and diastereoselectivity.
Table 1
Ligand Optimization
for the Reductive
Coupling Using 15aa
Entry
Ligand
%Yield 18ab
%Yield 19ab
% 9ab
1
dcpe
<5
58
20
2
PCy3
<5
86
<5
3
P(adam)3
<5
28
31
4
XPhos
<5
22
38
5
P(NMe2)3
<5
66
<5
6
P(OEt)3
<5
60
<5
7
(PhO)2PNMe2
<5
66
<5
8
SIMes
<5
8
67
9c
PCy3
<5
51
<5
10d
PCy3
<5
54
<5
11e
PCy3
<5
52
21
12f
PCy3
<5
31
14
13g
PCy3
90
<5
<5
129 mg (0.400 mmol) 9a, 96.6 mg (0.480 mmol) 15a, 5 mol % Cu(OAc)2, 6 mol % ligand, and 1.0
mL of toluene. A single diastereomer of
product was obtained in all cases by analysis of the unpurified reaction
mixture by 1H NMR spectroscopy. See the Supporting Information for further details.
Yield determined by 1H NMR
spectroscopy on the unpurified reaction mixture using dimethylfumarate
as analytical standard.
Reaction performed in MTBE.
Reaction performed in dioxane.
Reaction performed in CH2Cl2.
Reaction performed in THF.
Performed using 2.0 equiv of t-BuOH as additive. DMB = 2,4-dimethoxybenzyl.
Scheme 3
Proposed Reaction Catalytic Cycle
129 mg (0.400 mmol) 9a, 96.6 mg (0.480 mmol) 15a, 5 mol % Cu(OAc)2, 6 mol % ligand, and 1.0
mL of toluene. A single diastereomer of
product was obtained in all cases by analysis of the unpurified reaction
mixture by 1HNMR spectroscopy. See the Supporting Information for further details.Yield determined by 1HNMR
spectroscopy on the unpurified reaction mixture using dimethylfumarate
as analytical standard.Reaction performed in MTBE.Reaction performed in dioxane.Reaction performed in CH2Cl2.Reaction performed in THF.Performed using 2.0 equiv of t-BuOH as additive. DMB = 2,4-dimethoxybenzyl.An initial working hypothesis to
understand the difference in product
selectivity between the formation of urea19a in the
absence of t-BuOH versus the exclusive formation
of diamino-derivative 18a when t-BuOH
was used as an additive is given in Scheme . Regioselective hydrocupration of allenamide 15a by the LCuH[23,25c] catalyst 20 initially is expected to afford substituted linear allylcopper reagent 21 that may undergo E/Z isomerization
through σ–π–σ equilibration prior
to reaction with the imine electrophile. Then, diastereoselective
reaction of intermediate 21 with imine 9a provides Cu-amide intermediate 22. To afford product 18a from 22, direct silylation of the amine by
the silane must occur to regenerate the LCuHcatalyst 20; however, this step is expected to be slow due to the weak strength
of the N–Si bond (BDE ≈ 104 kcal/mol).[27] Due to the strong basicity of the N-anion
in 22, intramolecular attack of the oxazolidinonecarbonyl
may occur competitively to provide 23 containing an O–Cu
bond that should more easily silylate due to the high bond strength
of the O–Si bond (BDE ≈ 190 kcal/mol)[28] affording urea 24 and regenerating the LCuHcatalyst (20). Alternatively, when t-BuOH is present, protonation of the Cu–N bond of 22 by t-BuOH to afford product 18a directly
and generate LCuOBu is thermodynamically
favorable based on the pKa values for
a secondary amine (pyrrolidine: ∼44)[29] vs t-BuOH (32) and supported by DFT calculations
(vide infra).[30] The LCu-OBu intermediate can then undergo silylation
to regenerate the LCuHcatalyst 20. The role of alcohol
additives to facilitate catalyst turnover by protonation of Cu–N
intermediates has been documented previously.[25c,31] Sterically hindered alcohols such as t-BuOH have
been shown to be preferred since the rate of competitive protonation
of the Cu–H catalyst is reduced with bulky alcohols.[31]Next, the substrate scope of the Cu-catalyzed
reductive coupling
reaction using t-BuOH as additive to provide branched
diamino-derived products 18 was examined (Scheme ). In all cases, a single diastereomer
(the (S,S,S)-diastereomer) of product was obtained
as determined by analysis of the unpurified reaction mixture by 1HNMR spectroscopy. In general, a wide variety of iminescould
be employed in the reaction in good to excellent yields. Electron-deficient
(18a – 18k) and electron-rich (18l–18o) aryl groups both performed well
in the reaction. Heterocyclic imines (18k, 18s–18u) and C-substituted arenes
(18p–18r) were also well tolerated.
Finally, a sterically demanding imine (18m) or a m-NO2Ph group (18i) required heating
at 65 °C to afford good reactivity. Use of an aliphaticaldimine
(i.e., Ar1 = Me) did not provide any desired products.
Scheme 4
Imine Generality in the Cu-Catalyzed Reductive Coupling To Access
1,2-Diamino Synthons 18
Conditions: 9 (0.400
mmol), 15a (96.6 mg, 0.48 mmol), Cu(OAc)2 (5
mol %), PCy3 (6.5 mol %), t-BuOH (76 μL,
0.80 mmol), Me(MeO)2SiH (99 μL, 0.80 mmol), and 1.0
mL of toluene, rt 24 h followed by treatment with NH4F/MeOH.
See the Supporting Information for more
details. A single diastereomer of product was obtained in all cases
by analysis of the unpurified reaction mixture by 1H NMR
spectroscopy. Yields represent isolated yield.
Reaction performed at 65 °C.
Isolated as an inseparable mixture of 18 and urea 19.
Imine Generality in the Cu-Catalyzed Reductive Coupling To Access
1,2-Diamino Synthons 18
Conditions: 9 (0.400
mmol), 15a (96.6 mg, 0.48 mmol), Cu(OAc)2 (5
mol %), PCy3 (6.5 mol %), t-BuOH (76 μL,
0.80 mmol), Me(MeO)2SiH (99 μL, 0.80 mmol), and 1.0
mL of toluene, rt 24 h followed by treatment with NH4F/MeOH.
See the Supporting Information for more
details. A single diastereomer of product was obtained in all cases
by analysis of the unpurified reaction mixture by 1HNMR
spectroscopy. Yields represent isolated yield.Reaction performed at 65 °C.Isolated as an inseparable mixture of 18 and urea 19.Initial
analysis of the substrate scope for the urea-forming Cu-catalyzed
reductive coupling reaction employing DMB-substituted imines in the
absence of t-BuOH proved to be less general than
the analogous reaction conducted with t-BuOH as the
additive. In these problematiccases, a poor yield of desired product
was obtained even at 65 °C; however, the imine remained while
the allenamide had been consumed. As a result, the effect of the N-substituent of the imine electrophile was examined to
improve the efficiency of the reaction to the desired product (Table ). As an example,
the 2-naphthyl N-DMB-imine (9pa) afforded
a poor yield in the desired reaction (entry 1). A strong influence
on reactivity and the electroniccharacter of the aryl group (Ar2) of the imine was found (entries 1–5). Use of an electron-poor
aryl group (entry 5) afforded the best reaction yield; however, a
simple benzyl group also provided good reactivity (entry 3). As a
result, for problematicDMB-derived imines, the reactivity can be
improved by utilizing PMB, Bn, or p-CF3-benzyl as the N-substituent on the aldimine.
Table 2
Effect of Imine N-Substitution on
Reactivitya
Entry
Ar2
% yieldb
1
2,4-dimethoxyphenyl (9pa)
17 (19pa)
2
4-methoxyphenyl (9pb)
58 (19pb)
3
Phenyl (9pc)
70 (19pc)
4
4-fluorophenyl (9pd)
71
(19pd)
5
4-trifluoromethylphenyl
(9pe)
79 (19pe)
Conditions: 9p (0.400
mmol), 106 mg (0.480 mmol) of 15a, 5 mol % Cu(OAc)2, 6 mol % PCy3, 99 μL (0.80 mmol) of (MeO)2MeSiH, and 1.0 mL of toluene. A single diastereomer of product
was obtained in all cases by analysis of the unpurified reaction mixture
by 1H NMR spectroscopy.
Yield determined by 1H NMR spectroscopy on the unpurified
reaction mixture using dimethylfumarate
as analytical standard.
Conditions: 9p (0.400
mmol), 106 mg (0.480 mmol) of 15a, 5 mol % Cu(OAc)2, 6 mol % PCy3, 99 μL (0.80 mmol) of (MeO)2MeSiH, and 1.0 mL of toluene. A single diastereomer of product
was obtained in all cases by analysis of the unpurified reaction mixture
by 1HNMR spectroscopy.Yield determined by 1HNMR spectroscopy on the unpurified
reaction mixture using dimethylfumarate
as analytical standard.Based on the results from Table , the substrate scope for the urea-forming Cu-catalyzed
reductive coupling reaction in the absence of t-BuOH
was investigated using this new knowledge (Table ). DMB-substituted iminescould be employed
in good yields affording single diastereomers of product at room temperature
when Ar1 was a simple phenyl group (19b),
heterocyclic (19k, 19s, 19t), or substituted at the para-position with an electron-donating
group (19l, 19n) or an electron-withdrawing
group (19a). However, reactions employing iminescontaining
halogenated arenes or more sterically demanding aryl groups were
not successful utilizing the N-DMB derived imine
and instead required heating and the use of either an N-Bn or an N-CH2-p-CF3Ph group on the aldimine (see 19de, 19ee, 19j and 19m, 19pe, respectively).
Table 3
Imine Generality in the Cu-Catalyzed
Reductive Coupling To Access Chiral Ureasa
Conditions: 9 (0.40
mmol), 15a (96.6 mg, 0.48 mmol), Cu(OAc)2 (5
mol %), PCy3 (6.5 mol %), Me(MeO)2SiH (99 μL,
0.80 mmol), and 1.0 mL of toluene, rt 24 h followed by treatment with
NH4F/MeOH. See the Supporting Information for more details. A single diastereomer of product was obtained
in all cases by analysis of the unpurified reaction mixture by 1H NMR spectroscopy. Yields represent isolated yield.
Reaction performed at 65 °C.
Isolated as an inseparable
mixture
of urea 19 and 18.
Conditions: 9 (0.40
mmol), 15a (96.6 mg, 0.48 mmol), Cu(OAc)2 (5
mol %), PCy3 (6.5 mol %), Me(MeO)2SiH (99 μL,
0.80 mmol), and 1.0 mL of toluene, rt 24 h followed by treatment with
NH4F/MeOH. See the Supporting Information for more details. A single diastereomer of product was obtained
in all cases by analysis of the unpurified reaction mixture by 1HNMR spectroscopy. Yields represent isolated yield.Reaction performed at 65 °C.Isolated as an inseparable
mixture
of urea 19 and 18.Stereochemical assignment of the products obtained
in the Cu-catalyzed
reductive coupling reaction as the (S,S,S)-diastereomer
was determined unequivocally by X-ray crystallography (Scheme ). While the branched products 18 were typically noncrystalline, formation of the HCl-salt
of 18c afforded crystalline material whose structure
was determined by single-crystal Xray analysis. Furthermore, conversion
of products 18 to the urea 19 could also
be achieved after isolation of 18 by subsequent treatment
with n-BuLi (e.g., 18a → 19a). The urea product obtained from this sequence was identical
to the material made from the reductive coupling reaction performed
in the absence of t-BuOH by NMR spectroscopy confirming
that the same stereoisomer of product was formed in both reductive
coupling processes (i.e., with or without t-BuOH
as additive).
Scheme 5
Stereochemistry Determination
The synthetic utility of the reaction products obtained in the
allenamide/imine reductive coupling reaction is highlighted in Scheme . The phenethyl group
of urea19a derived from the Evans oxazolidinone of the
allenamide starting material could be cleaved in a three-step sequence
consisting of alcohol activation (MsCl), base induced elimination,
and enamide hydrolysis with aqueous acid to provideurea 25 in good overall yield without isolation of intermediates. Furthermore,
synthon 27 is a viable intermediate to access chiral
aminopiperidine 28 for the preparation of the potent
NK-1 inhibitor compounds CP-99,994 and CP-122,721,[2a,7] which
could easily be accessed from reductive coupling product 18c (Scheme ). The Cu-catalyzed
reductive coupling was scaled to 1.0 g without the need for an inert
atmosphere glovebox by performing the reaction on the “benchtop”
using standard Schlenk techniques and preparing the (PCy3)Cu-catalyst by adding the PCy3 as a 20 wt % solution
in toluene that is commercially available.[32] The catalyst loading could be reduced to 2.0 mol % Cu providing 18c in good yield and excellent diastereocontrol in only 2
h of reaction time. Considering the low cost and high availability
of the Cu-precatalyst,[32] the ligand employed
(PCy3),[32] and the chiral allenamide 15a,[26] along with the high catalytic
activity of this system (2 mol % catalyst loading), the current method
represents a highly practical and scalable method for the synthesis
of diamino-synthons 18/19. Allylation of 18c was then carried out using allyl bromide, followed by
ring-closing metathesis with the Hoveyda–Grubbs second generation
catalyst to provide access to compound 27 as an orthogonally
protected aminopiperidine derivative as a single stereoisomer.
Scheme 6
Synthetic Applications
Mechanistic Modeling by DFT Analysis
To shed light
onto the mechanism and origin of diastereoselectivity, we used dispersion-corrected
DFT calculations (see Supporting Information for details). Specifically, we performed extensive conformational
analysis on all intermediates and transition states using the B3LYP-D3
functional and a def2-SVP basis set[33] with
toluene as the solvent using the CPCM solvation model[34] as implemented in Gaussian16. Further, to refine the energetics
and compare methods, single-point calculations using the M06-L functional,[35] as well as a larger basis set (def2-TZVPP) with
B3LYP-D3, which yielded similar energetic profiles, were subsequently
performed. For simplicity, only B3LYP-D3/def2-SVP optimization energetics
will be discussed in the text. Structures were visualized using CYLview
Version 1.0.561.[36]As shown in Figure , initial investigations
were conducted by first analyzing the hydrocupration of allenamide 15a with (PCy3)CuH as catalyst. Following coordination
of the copper and allenamide π-bond, the energetically favored
hydrocupration proceeds via TS–I-II (barrier of
10.4 kcal/mol with respect to separated 15a and LCuH structures) to form the branched allylcopper species II. Presumably this transition state benefits from lack of
steric hindrance between the ligand and the chiral auxiliary, which
were present in the alternative transition states. Specifically, alternate
hydrocupration transition states leading to linear allylcopper species
(TS–I-III and TS–I-III) were
found to be much higher in energy by ∼3 kcal/mol for TS–I-III and by
>7 kcal/mol for all other pathways and were therefore not productive.
Figure 2
Structures
and relative free energies (in kcal/mol, with respect
to separate LCuH catalyst and reactants) of possible
hydrocupration pathways, optimized using B3LYP-D3/def2SVP-CPCM(toluene),
M06-L/def2SVP-gas//B3LYP-D3/def2SVP-CPCM(toluene) (in parentheses),
and B3LYP-D3/def2TZVPP-gas//B3LYP-D3/def2SVP-CPCM(toluene) {in braces}.
Structures
and relative free energies (in kcal/mol, with respect
to separate LCuHcatalyst and reactants) of possible
hydrocupration pathways, optimized using B3LYP-D3/def2SVP-CPCM(toluene),
M06-L/def2SVP-gas//B3LYP-D3/def2SVP-CPCM(toluene) (in parentheses),
and B3LYP-D3/def2TZVPP-gas//B3LYP-D3/def2SVP-CPCM(toluene) {in braces}.In turn, the branched allylcopper intermediate
is expected to undergo
isomerization to linear allylcopper species by σ–π–σ
isomerization. Recently, Buchwald and co-workers reported branched-linear
allylcopper isomerizations for a system with a bidentate phosphine
ligand[25c] as well as with a CuH-catalyzed
allylation of ketones and dienes.[37] In
our case, it was calculated that the branched allylcopper intermediate II can readily isomerize (barrier of only 6.9 and 10.0 kcal/mol
via TS–II-IIIor TS–II-III, respectively) to form the nearly isoenergeticcis or trans linear allylcopper intermediates (III and III). Intermediate III was slightly favorable compared
to intermediate III (by 0.6 kcal/mol), as was the cis isomerization
transition state (TS–I-III was favored by 3.1 kcal/mol), presumably due to coordination
between the oxazolidinone and the copper (Cu–O bond distance
= 2.37 Å). However, upon coordination of the imine to the copper,
the trans conformation (III′) becomes significantly more favored (as
seen in Figure ),
likely due to unfavorable steric hindrance between the imine and the
oxazolidinone in the III′ conformation (see Supporting Information for further details).
Figure 3
Structures and relative free energies (in kcal/mol,
with respect
to separate LCuH catalyst and reactants) for proposed
mechanistic pathway, optimized using B3LYP-D3/def2SVP-CPCM(toluene),
M06-L/def2SVP-gas//B3LYP-D3/def2SVP-CPCM(toluene) (in parentheses),
and B3LYP-D3/def2TZVPP-gas//B3LYP-D3/def2SVP-CPCM(toluene) {in braces}.
Optimized structures of transition states visualized with CYLview
are shown (with PCy3 ligand faded out for clarity).
Structures and relative free energies (in kcal/mol,
with respect
to separate LCuHcatalyst and reactants) for proposed
mechanistic pathway, optimized using B3LYP-D3/def2SVP-CPCM(toluene),
M06-L/def2SVP-gas//B3LYP-D3/def2SVP-CPCM(toluene) (in parentheses),
and B3LYP-D3/def2TZVPP-gas//B3LYP-D3/def2SVP-CPCM(toluene) {in braces}.
Optimized structures of transition states visualized with CYLview
are shown (with PCy3 ligand faded out for clarity).Next, we focused on the key C–C bond formation
steps. As
shown in Figure ,
after performing extensive conformational analysis on the subsequent
diastereomericC–C bond forming transition states with the
allylcopper intermediates and the imine substrate (see Supporting Information for details), the most
favorable pathway for diastereoselective C–C bond formation
was identified to proceed from the trans linear intermediate III′ through a
Zimmerman–Traxler transition state TS-III′-IV (barrier of only 7.4
kcal/mol from complexed III′ intermediate) to branched addition product IV. Further, in agreement with
experiment, the competing diastereomeric transition state TS-III′-IV which would lead to
the opposite diastereomer was determined to be much higher in energy.
Notably, all C–C bond formation steps fromare reversible, as the branched addition products IV and IV were each uphill in energy
(by ∼2 kcal/mol and ∼6 kcal/mol respectively), which
can have implications for rational catalyst and reaction design (vide infra).To gain insights into the origin of diastereoselectivity,
we performed
distortion–interaction and NCI analysis (Figure ). Overall, comparing the structures of the
lowest energy competing diastereomeric transition states TS-III′-IV() and TS-III′-IV() reveals that the structures of
these transition states were remarkably similar, with key C–C
and C–Cu bond distances differing by no more than 0.05 Å.
However, the orientation of the chiral auxiliary is different, as
the TS-III′-IV() transition
state has the oxazolidinone moiety of the enamide group of the substituted
Cu(allyl) ligand in an s-trans conformation while
the TS-III′-IV() has
this group in an s-cis conformation that, as shown
in Figure , leads
to a 2.2 kcal/mol energy destabilization. Furthermore, the ground
state structures of chiral oxazolidinone-derived enamides are known
to favor an s-trans conformation.[38] In addition, distortion–interaction analysis[39] (Figure a) showed that the distortion energy of the TS-III′-IV() transition state was higher than
that of the corresponding TS-III′-IV() transition state (by 3.9 kcal/mol). However, the (S,S,S) system benefited from much stronger interaction energy
(by 8.5 kcal/mol). Overall, this favorable interaction between the
imine and allylcopper makes the TS-III′-IV() the favorable diastereomeric transition state.
Finally, noncovalent interaction (NCI) analysis (performed using Multiwfn[40] software and visualized using VMD[41] software) further supports the presence of favorable
interactions in the TS-III′-IV() transition state (Figure b). Specifically, in both transition states, there
appeared to be favorable C–H···π interactions
between the ligand and the benzyl group of the imine (highlighted
inside the blue circle). However, comparing the areas in red circles,
the TS-III′-IV() system
had stronger noncovalent interactions between the oxazolidinone group
and the phenyl ring on the imine. This suggests that noncovalent interactions
(i.e., between the oxazolidinone moiety and the protecting group)
are critical for control of diastereoselectivity. Taken together,
these results suggest that both the conformational preference for
the s-trans geometry about the N-enamide group of the substituted Cu(allyl) ligand and favorable
noncovalent interactions between the oxazolidinone group and the imine
are the major contributing factors for diastereocontrol in these reactions.
Figure 4
(A) Distortion–Interaction
analysis of key diastereomeric
C–C bond formation transition states. Electronic energies reported
at B3LYP-D3/def2SVP-CPCM(toluene) level of theory. (B) Noncovalent
Interaction analysis of key diastereomeric C–C bond formation
transition states. Color code for the atoms is shown.
Figure 5
Energetic comparison of s-trans and s-cis conformations of an (E)-enamide system
with chiral oxazolidinone, with steric hindrance causing allylic strain
highlighted. Structures optimized using B3LYP-D3/def2SVP-CPCM(toluene)
(Hrel and Grel shown in kcal/mol).
(A) Distortion–Interaction
analysis of key diastereomericC–C bond formation transition states. Electronic energies reported
at B3LYP-D3/def2SVP-CPCM(toluene) level of theory. (B) Noncovalent
Interaction analysis of key diastereomericC–C bond formation
transition states. Color code for the atoms is shown.Energeticcomparison of s-trans and s-cis conformations of an (E)-enamide system
with chiral oxazolidinone, with steric hindrance causing allylic strain
highlighted. Structures optimized using B3LYP-D3/def2SVP-CPCM(toluene)
(Hrel and Grel shown in kcal/mol).Following C–C bond formation, intermediate IV serves as a fork between two reaction pathways depending on whether
or not there is t-BuOH present (as supported by experiments; vide supra). Specifically, in the presence of t-BuOH, the alcoholcan act as a proton source to protonate the Cu–N
bond and yield branched product 18, while the alkoxide
binds to the copper. From this, the silane reagent can exchange hydride
for the alkoxide group, reforming the catalyst. While this protonation
of the amine moiety and concomitant release of the t-BuO-CuL was calculated to be initially energetically unfavorable
(uphill by ∼4 kcal/mol), the exchange of hydride for the alkoxide
was thermodynamically favorable (downhill by ∼10 kcal/mol),
rendering this overall process energetically feasible. In the absence
of t-BuOH, a thermodynamically favorable rearrangement
of intermediate IV can yield the Cu-alkoxideurea intermediate V (∼7 kcal/mol exergonic), which can then readily undergo
transmetalation with silane to reform the LCuHcatalyst and furnish
the silylated product of urea 19. This mechanistic model
is consistent with experimental findings that the presence of t-BuOH has a profound effect on the product selectivity
(but not diastereoselectivity) of the reaction toward either of the
products (vide supra).As previously noted,
computational modeling of the imine addition
predicts this step should be reversible. This phenomenon has important
impacts for future developments of catalyst controlled enantioselective
reactions utilizing a chiral catalyst in conjunction with an achiral
allenamide. In this regard, reaction of achiral allenamide 15b with imine 9a using (S,S)-Ph-BPE as a chiral ligand was examined with and without t-BuOH as the additive (Scheme ). Again, branched product 29 was formed as a single diastereomer when t-BuOH
was present in the reaction, and urea 30 was formed as
a single diastereomer in the absence of t-BuOH. Separate
conversion of 29 to 30 using n-BuLi confirmed that the same relative stereochemistry was formed
in both reactions. Importantly, 29 and 30 were formed in different enantiopurities (57:43 vs 80:20 er, respectively), supporting a reversible imine addition step in these reactions. For example, if imine addition were irreversible, reaction of 15b with a chiral catalyst to afford the analogous intermediate
to 22 (Scheme ) must be enantiodetermining and requires that urea product 30 formed from rearrangement of the intermediate 22 derivative to have identical enantiopurity to that of 29. However, when a chiral ligand is employed, rearrangement of the
two enantiomers of intermediate 22 may occur at different
rates because the transition states will be diastereomeric due to
the chirality on the ligand. Therefore, the carbamate migration step
may also be enantiodetermining if the imine addition step becomes
reversible enabling different enantiomeric ratios to be obtained for
a 29-selective vs a 30-selective process
as was observed.
Scheme 7
Mechanistic Implications Relevant to Catalyst-Controlled
Enantioinduction
Conclusion
In
conclusion, a highly stereoselective method for the reductive
coupling of imines with a chiral allenamide was developed as a convenient
strategy for the asymmetric synthesis of valuable 1,2-diamino synthons.
The method employs readily available and cost-effective starting materials[26] and catalyst (Cu(OAc)2/PCy3)[32] and can be performed on the “bench-top”
using standard Schlenk techniques without issues. Use of tert-butanol as an additive was shown to aid in the amine release and
catalyst regeneration to avoid the formation of urea products that
are exclusively obtained in the absence of this additive. The oxazolidinone
moiety of the final products could be removed chemoselectively without
disruption of the pendant terminal alkene, and an orthogonally protected
chiral aminopiperidine derivative en route to important biologically
active pharmaceuticals was demonstrated. Finally, mechanistic investigations
by density functional theory calculations identified the mechanism
for stereoselection in these processes as determined from the relative
transition state barriers of N-substituted allylcoppercomplexes to the imine electrophile. This C–C bond forming
addition step was shown to be reversible by calculation and was experimentally
supported by the catalytic asymmetric reaction of a chiral catalyst
with an achiral allenamide. These mechanistic insights are important
for the development of future asymmetriccatalyst-controlled procedures
and are currently under further investigation in these laboratories.
Experimental Section
General
1HNMR spectra were recorded on
Bruker 600 MHz spectrometers. Chemical shifts are reported in ppm
from tetramethylsilane with the solvent resonance as the internal
standard (CDCl3: 7.26 ppm). Data are reported as follows:
chemical shift, integration, multiplicity (s = singlet, d = doublet,
t = triplet, q = quartet, p = pentet, h = hextet, hept = heptet, br
= broad, m = multiplet), and coupling constants (Hz). 13CNMR spectra were recorded on a Bruker 600 MHz (151 MHz) instrument
with complete proton decoupling. Chemical shifts are reported in ppm
from tetramethylsilane with the solvent as the internal standard (CDCl3: 77.0 ppm). Liquid chromatography was performed using forced
flow (flash chromatography) on silica gel purchased from Silicycle.
Thin layer chromatography (TLC) was performed on glass-backed 250
μm silica gel F254 plates purchased from Silicycle.
Visualization was achieved by using UV light, a 10% solution of phosphomolybdic
acid in EtOH, or potassium permanganate in water followed by heating.
HRMS was collected using a Jeol AccuTOF-DART mass spectrometer using
DART source ionization. All reactions were conducted in oven or flame-dried
glassware under an inert atmosphere of nitrogen or argon with magnetic
stirring unless otherwise noted. Solvents were obtained from VWR as
HPLC grade and transferred to septa sealed bottles, degassed by argon
sparge, and analyzed by Karl Fischer titration to ensure watercontent
was ≤600 ppm. Me(MeO)2SiH was purchased from Alfa
Aesar and used as received. Allenamides 15 were prepared
in one step as described in the literature.[26] Aldehydes were purchased from Sigma-Aldrich, Combi-Blocks, TCI America,
Alfa Aesar, or Oakwood Chemicals and used as received. Tricyclohexylphosphine
and Cu(OAc)2 were purchased from the Strem Chemical Company
and used as received. All other materials were purchased from VWR,
Sigma-Aldrich, Combi-Blocks, or Alfa Aesar and used as received. Imines 9a,[42]9b,[43]9c,[44]9ee,[45]9h,[46]9i,[47]9j,[48]9l,[45]9m,[47]9pb,[49]9pc,[50] and 9u(45) were synthesized as described in the literature.
General Procedure
A for the Synthesis of Imines
A 25
mL round-bottom flask equipped with a magnetic stirring bar was charged
with aldehyde (6.0 mmol, 1.0 equiv) and dichloromethane (8 mL). Anhydrous
magnesium sulfate was added to this solution while stirring followed
by 2,4-dimethoxy benzylamine (6.0 mmol, 1.0 equiv) dropwise. The reaction
mixture was stirred at room temperature for 12 h under a nitrogen
atmosphere. After the reaction is complete the crude reaction mixture
was filtered through Celite to remove magnesium sulfate. The filtrate
was concentrated in vacuo to yield the pure imine,
which was stored under nitrogen in the fridge.
Following General Procedure A, thiophene-2-carboxaldehyde
(0.67 g, 6.0 mmol), 2,4-dimethoxybenzylamine (1.0 g, 6.0 mmol),
magnesium sulfate (2.0 g), and dichloromethane (8 mL) were used. The
title compound was obtained as a yellow solid (1.15 g, 74%). Mp −47.4–50.2
°C. 1HNMR (600 MHz, CDCl3) δ 8.36
(s, 1H), 7.37 (d, J = 6 Hz, 1H), 7.29 (d, J = 3.6 Hz, 1H), 7.18 (d, J = 8.4 Hz, 1H),
7.06 (t, J = 6 Hz, 1H), 6.50–6.46 (m, 2H),
4.74 (s, 2H), 3.80 (s, 6H). 13C{1H} NMR (151 MHz, CDCl3) δ 160.2, 158.3, 154.8, 142.9, 130.3, 130.2, 128.6,
127.2, 119.6, 104.1, 98.4, 58.2, 55.4, 55.3. HRMS (DART) m/z calcd for C14H16NO2S [M + H]+: 262.0902; Found [M + H]+: 262.0915.
General Procedure B for the Synthesis of 18
To a 20 mL crimp cap vial with a stir bar in an
Ar filled glovebox
were charged Cu(OAc)2 (3.6 mg, 20 μmol) and PCy3 (7.3 mg, 26 μmol) followed by toluene (1.0 mL) and tert-butanol (76.5 μL, 2 equiv). The mixture was stirred
for 5 min. Allenamide 15a (96.6 mg, 480 μmol) followed
by imine (400 μmol) was then charged, and the vial was sealed
with a crimp-cap septum and removed from the glovebox. Dimethoxymethylsilane
(0.099 mL, 2 equiv) was then charged to the reaction mixture (. The mixture was then stirred at rt for 24 h. The reaction was quenched
by addition of 200 mg of NH4F and 2.5 mL of MeOH followed
by agitation at rt for 30 min. A 10 mL volumen of 5% NaHCO3 was then added to the mixture followed by extraction with DCM (2
× 5 mL). The combined organics were dried with Na2SO4, filtered, and concentrated in vacuo. Crude product
was purified by flash chromatography on silica gel to afford the desired
product.
According to General Procedure B, the
product was purified by silica gelchromatography (3% E.A. in DCM)
to provide 172 mg (79%) of 18u as a pale-yellow foam
as a single diastereomer and a 95:5 mixture of the branched to rearranged
product. R = 0.38 (50%
EtOAc/hexanes). 1HNMR (600 MHz, CDCl3) δ
7.35 (dd, J = 4.2, 2.5 Hz, 3H), 7.17 (dd, J = 6.0, 2.6 Hz, 2H), 7.04 (d, J = 8.1
Hz, 1H), 6.84 (d, J = 6 Hz, 1H), 6.59 (d, J = 6 Hz, 1H), 6.50 (s, 1H), 6.48 (d, J = 12 Hz, 1H), 5.24 (dt, J = 16.7, 9.5 Hz, 1H),
4.89–4.83 (m, 2H), 4.54 (t, J = 8.0 Hz, 1H),
4.50 (t, J = 12 Hz, 1H), 4.39 (d, J = 9.6 Hz, 1H), 4.11 (t, J = 12 Hz, 1H), 3.84 (s,
3H), 3.83 (s, 3H), 3.80 (d, J = 13.5 Hz, 1H), 3.51
(d, J = 13.4 Hz, 1H), 2.24 (s, 1H). 13C{1H} NMR (151 MHz, CDCl3) δ 160.3, 158.8, 158.0,
147.7, 138.0, 132.2, 130.7, 129.2, 129.1, 129.1, 127.7, 126.4, 120.3,
119.8, 111.6, 103.6, 98.6, 70.2, 63.6, 59.4, 57.7, 55.4, 55.2, 46.1.
HRMS (DART) m/z calcd for C26H28BrN2O4S [M + H]+: 543.0953; Found [M + H]+: 543.0949.
General Procedure
C for the Synthesis of 19
To a 20 mL crimp cap
vial with a stir bar in an Ar filled glovebox
were charged Cu(OAc)2 (3.6 mg, 20 μmol) and PCy3 (7.3 mg, 26 μmol) followed by toluene (1.0 mL), and
the mixture was stirred for 5 min. Allenamide 15a (96.6
mg, 480 μmol) followed by imine (400 μmol) was then charged,
and the vial was sealed with a crimp-cap septum and removed from the
glovebox. Dimethoxymethylsilane (0.099 mL, 2 equiv) was
then charged to the reaction mixture (. The mixture
was then stirred at rt for 24 h. The reaction was quenched by addition
of 200 mg of NH4F and 2.5 mL of MeOH followed by agitation
at rt for 30 min. A 10 mL volume of 5% NaHCO3 was then
added to the mixture followed by extraction with DCM (2 × 5 mL).
The combined organics were dried with Na2SO4, filtered, and concentrated in vacuo. Crude product was purified
by flash chromatography on silica gel to afford the desired product 19.
The reaction was set up according to
general procedure C and stirred at 65 °C for 24 h. The product
was purified by silica gelchromatography (5% E.A. in DCM) to provide
132 mg (71%) of 19pd as a colorless foam as a single
diastereomer. R = 0.62
(20% EtOAc/DCM). 1HNMR (600 MHz, CDCl3) δ
7.85–7.75 (m, 3H), 7.54–7.48 (m, 3H), 7.36 (t, J = 7.6 Hz, 2H), 7.33–7.26 (m, 3H), 7.25–7.22
(m, 1H), 7.09 (dd, J = 8.4, 5.5 Hz, 2H), 6.98 (t, J = 8.6 Hz, 2H), 5.69 (ddd, J = 17.0, 10.2,
8.6 Hz, 1H), 5.20 (d, J = 10.1 Hz, 1H), 4.96 (d, J = 14.9 Hz, 1H), 4.91 (t, J = 6.1 Hz,
1H), 4.86 (d, J = 17.0 Hz, 1H), 4.39–4.30
(m, 2H), 4.15–4.08 (m, 2H), 3.68–3.61 (m, 2H). 13C{1H} NMR (151 MHz, CDCl3) δ 163.12 (C–F, 1J C–F = 246.13 Hz), 161.49 (C–F, 1J C–F = 246.13 Hz), 160.8, 137.6,
134.8, 134.7, 133.3, 133.1, 130.46, 132.02 (C–F, 3J C–F = 4.53 Hz), 131.99 (C–F, 3J C–F = 4.53 Hz), 130.46, 130.41,
129.1, 128.7, 127.87, 127.80, 127.7, 127.1, 126.6, 126.5, 124.3, 121.4,
115.60 (C–F, 2J C–F = 21.14
Hz), 115.46 (C–F, 2J C–F
= 21.14 Hz), 66.3, 64.9, 63.6, 62.0, 45.0. 19FNMR (565
MHz, CDCl3) δ – 114.61. HRMS (DART) m/z calcd for C30H28FN2O2 [M + H]+: 467.2135; Found
[M + H]+: 467.2105.
Synthesis of 29 from Achiral Allenamide 15b
To a 20 mL crimp
cap vial with a stir bar in an Ar filled
glovebox were charged Cu(OAc)2 (1.8 mg, 10 μmol)
and Ph-BPE (5.1 mg, 10 μmol) followed by toluene (2.5 mL) and tert-butanol (26.3 μL, 275 μmol). The mixture
was stirred for 5 min. Allenamide 15b (37.5 mg, 300 μmol)
followed by imine 9a (250 μmol) was then charged,
and the vial was sealed with a crimp-cap septum and removed from the
glovebox. Dimethoxymethylsilane (0.061 mL, 2 equiv) was charged
to the reaction mixture, and the reaction mixture was stirred at rt
for 24 h. The reaction was quenched by addition of 100 mg of NH4F and 1.5 mL of MeOH followed by agitation at rt for 30 min.
A 5 mL volume of 5% NaHCO3 was then added to the mixture
followed by extraction with DCM (2 × 3 mL). The combined organics
were dried with Na2SO4, filtered, and concentrated in vacuo. Crude product was purified by flash chromatography
on silica gel (25% EtOAc/hexanes) to afford 78 mg (69%) of 29 as a white solid as a single diastereomer and as a 57:43 mixture
of enantiomers as determined via chiral HPLC analysis (Chiracel AD-3
85:15 heptane/isopropanol 1.50 mL/min, 254 nm). R = 0.45 (50% EtOAc/hexanes). 1HNMR (600 MHz, CDCl3) δ 7.60 (d, J = 7.8 Hz, 2H), 7.47 (d, J = 7.9 Hz, 2H), 6.91 (d, J = 8.2 Hz, 1H), 6.44 (d, J = 2.3 Hz, 1H),
6.40 (dd, J = 8.1, 2.4 Hz, 1H), 5.47 (ddd, J = 17.2, 10.5, 6.9 Hz, 1H), 5.05 (d, J = 10.6 Hz, 1H), 4.99 (d, J = 17.2 Hz, 1H), 4.46
(t, J = 6.0 Hz, 1H), 4.29 (q, J =
12.0 Hz, 1H), 4.21 (q, J = 12.0 Hz, 1H), 3.79 (s,
6H), 3.75 (d, J = 9.6 Hz, 1H), 3.71 (d, J = 13.7 Hz, 1H), 3.42 (q, J = 8.1 Hz, 1H), 3.25
(d, J = 13.7 Hz, 1H), 3.16 (td, J = 8.7, 6.4 Hz, 1H). 13C{1H} NMR (151 MHz, CDCl3) δ 160.3, 158.7, 158.6, 144.8, 131.4, 130.7, 130.6, 130.2
(q, J = 31.71 Hz), 128.9, 128.7, 125.3 (q, J = 3.02 Hz), 119.7, 103.6, 98.6, 98.6, 62.2, 62.1, 61.7,
60.9, 55.3, 55.2, 46.1, 40.8. 19FNMR (565 MHz, CDCl3) δ −62.36. HRMS (DART) m/z calcd for C23H26F3N2O4 [M + H]+: 451.1845; Found [M + H]+: 451.1881.
Synthesis of 30 from Achiral
Allenamide 15b
To a 20 mL crimp cap vial with
a stir bar in an Ar filled
glovebox were charged Cu(OAc)2 (1.8 mg, 10 μmol)
and Ph-BPE (5.1 mg, 10 μmol) followed by toluene (0.5 mL). The
mixture was stirred for 5 min. Allenamide 15b (37.5 mg,
300 μmol) followed by imine 9a (250 μmol)
was then charged, and the vial was sealed with a crimp-cap septum
and removed from the glovebox. Dimethoxymethylsilane (0.061
mL, 2 equiv) was then charged to the reaction mixture. The mixture
was then stirred at rt for 24 h. The reaction was quenched by addition
of 100 mg of NH4F and 1.5 mL of MeOH followed by agitation
at rt for 30 min. A 5 mL volume of 5% NaHCO3 was then added
to the mixture followed by extraction with DCM (2 × 3 mL). The
combined organics were dried with Na2SO4, filtered,
and concentrated in vacuo. Crude product was purified by flash chromatography
on silica gel (50% EtOAc/hexanes) to afford 68 mg (60%) of 30 as a colorless liquid and as an 80:20 mixture of enantiomers as
determined via chiral HPLC analysis (Chiracel AD-3 90:10 heptane/isopropanol
1.00 mL/min, 220 nm). R = 0.28 (50% EtOAc/hexanes). 1HNMR (600 MHz, CDCl3) δ 7.59 (d, J = 8.0 Hz, 2H), 7.28
(d, J = 8.0 Hz, 2H), 7.00 (d, J =
8.3 Hz, 1H), 6.37 (dd, J = 8.3, 2.4 Hz, 1H), 6.32
(d, J = 2.4 Hz, 1H), 5.62 (ddd, J = 17.1, 10.1, 8.6 Hz, 1H), 5.22 (d, J = 10.1 Hz,
1H), 5.04 (d, J = 17.0 Hz, 1H), 4.70 (d, J = 14.7 Hz, 1H), 4.01 (d, J = 8.2 Hz,
1H), 3.86 (d, J = 14.7 Hz, 1H), 3.78 (s, 3H), 3.76–3.68
(m, 2H), 3.63 (t, J = 8.4 Hz, 1H), 3.55 (s, 3H),
3.36–3.31 (m, 1H), 3.26–3.22 (m, 1H). 13C{1H}
NMR (151 MHz, CDCl3) δ 162.3, 160.6, 158.6, 142.7,
135.0, 131.7, 130.6 (q, J = 31.71 Hz), 127.6, 126.7
(q, J = 273.31 Hz), 125.5 (q, J =
3.02 Hz), 121.2, 116.2, 104.1, 98.0, 68.2, 63.8, 62.2, 55.3, 54.8,
46.1, 40.8. 19FNMR (565 MHz, CDCl3) δ
−62.50. HRMS (DART) m/z calcd
for C23H26F3N2O4 [M + H]+: 451.1845; Found [M + H]+: 451.1882.
Synthesis of 18c on 1.0 g Scale
To a 20
mL crimp cap vial was charged Cu(OAc)2 (16.1 mg, 88.9 μmol),
and the vial was sealed with a crimp-cap septum. The vial was evacuated
and backfilled with nitrogen 3 times and then charged with toluene
(5 mL), 20% PCy3 solution in toluene (194 μL, 111
μmol), and tert-butanol (0.85 mL, 8.89 mmol),
and the mixture was allowed to stir at rt for 10 min until all the
Cu(OAc)2 dissolved. A 50 mL two-neck round-bottom flask
was then charged with imine 9c (1.0 g, 4.44 mmol) and
allene 15a (1.07 g, 5.33 mmol), and the flask was evacuated
and backfilled with nitrogen 3 times. The flask was then charged with
toluene (5 mL). The imine/allene flask was then charged with the catalyst
solution. Dimethoxymethyl silane (1.1 mL, 8.89 mmol) was charged to
the reaction mixture (, and the reaction was allowed to stir
at rt for 2 h. A 50 mL round-bottom flask was charged with NH4F (2 g) and MeOH (20 mL), and the reaction mixture was transferred
via pipet to this flask and allowed to stir at rt for 30 min. The
volatiles were concentrated in vacuo, and 50 mL of
5% NaHCO3 solution were added to the flask. The mixture
was extracted with CH2Cl2 (2 × 20 mL),
and the combined organics were dried over Na2SO4 and concentrated in vacuo. The crude residue was
purified by silica gelchromatography (5% EtOAc/DCM) to afford 1.655
g (91%) of 18c as a white solid as a single diastereomer.
Synthesis of 25
To a solution of 207.5
mg (393 μmol) of 19a in 2 mL of CH2Cl2 at 0 °C were charged 66 μL (473 μmol) of
triethylamine followed by dropwise addition of 30.5 μL (393
μmol) of MsCl. The mixture was stirred for 30 min at 0 °C,
and then 4 mL of 10% NH4Cl were added. The mixture was
extracted with CH2Cl2 (3 × 5 mL). The combined
organics were dried with anhydrous Na2SO4 and
filtered, and the volatiles were removed in vacuo. The crude residue was then dissolved in 2 mL of THF and cooled
to 0 °C. A 1.0 M concentration of potassium tert-butoxide (433 μL, 433 μmol) in THF was then added, and
the mixture was warmed to room temperature and stirred for 30 min.
To the mixture was added 5 mL of 10% brine followed by extraction
with CH2Cl2 (3 × 5 mL). The combined organics
were dried with anhydrous Na2SO4 and filtered,
and the volatiles were removed in vacuo. The crude
residue was then dissolved in 4 mL of THF in a crimp cap vial. To
the solution were then added 788 μL (3.94 mmol) of 5.0 M aqueous
H2SO4. The vial was purged with argon, sealed,
and immersed in an oil bath at 50 °C. After 3 h the reaction
mixture was cooled to room temperature, and 15 mL of saturated aqueous
NaHCO3 were added. The mixture was extracted with CH2Cl2 (3 × 5 mL). The combined organics were
dried with anhydrous Na2SO4 and filtered, and
the volatiles were removed in vacuo. The crude residue
was purified by flash chromatography (20% EtOAc/DCM) to afford 112
mg (70%) of 25 as a colorless foam. R = 0.22 (50% EtOAc/hexanes). 1HNMR (600 MHz, CDCl3) δ 7.59 (d, J = 8.0 Hz, 2H), 7.30 (d, J = 7.9 Hz, 2H), 7.02 (d, J = 8.3 Hz, 1H), 6.37 (d, J = 8.3 Hz, 1H),
6.31 (d, J = 2.4 Hz, 1H), 5.74 (ddd, J = 17.2, 10.2, 7.2 Hz, 1H), 5.41 (s, 1H), 5.12 (d, J = 10.2 Hz, 1H), 5.10 (d, J = 12 Hz, 1H), 4.67 (d, J = 14.8 Hz, 1H), 4.06 (d, J = 7.5 Hz,
1H), 3.86–3.82 (m, 2H), 3.76 (s, 3H), 3.54 (s, 3H). 13C{1H} NMR (151 MHz, CDCl3) δ 161.6, 160.5, 158.5,
143.2, 136.1, 131.4, 130.6 (q, J = 33.22 Hz), 127.5,
126.7 (q, J = 271.8 Hz), 125.5 (q, J = 3.02 Hz), 118.0, 116.5, 104.1, 97.9, 65.5, 62.1, 55.3, 54.9, 39.8. 19FNMR (565 MHz, CDCl3) δ −62.49.
HRMS (DART) m/z calcd for C21H22F3N2O3 [M
+ H]+: 407.1583; Found [M + H]+: 407.1600.
Synthesis of 19a from 18a
To a
solution of 100 mg (190 μmol) of 18a in 1.0
mL of THF at −10 °C were added 114 μL (285 μmol)
of 2.5 M solution of BuLi in hexanes.
The reaction mixture was allowed to warm to room temperature and stirred
for 1 h. To the mixture were added 2 mL of saturated NH4Cl, and the mixture was extracted with CH2Cl2 (3 × 3 mL). The combined organics were dried with anhydrous
Na2SO4 and filtered, and the volatiles were
removed in vacuo. The crude residue was purified
by flash chromatography (5% EtOAc/DCM) to afford 94 mg (94%) of 19a as a colorless foam.
Synthesis of 26
A crimp cap vial was charged
with 100 mg (233 μmol) of 18c, CH3CN
(1 mL), K2CO3 (161 mg, 1.17 mmol), TBAI (17.2
mg, 46.7 μmol), and allyl bromide (101 μL, 1.17 mmol).
The mixture was heated at 85 °C for 18 h. The reaction was quenched
with 5 mL of water, and the mixture was extracted with MTBE (3 ×
3 mL). The combined organics were dried with anhydrous Na2SO4 and filtered, and the volatiles were removed in vacuo. The crude residue was purified by flash chromatography
(20% EtOAc/hexanes) to afford 90 mg (82%) of 26 as a
yellow solid. R = 0.60
(50% EtOAc/hexanes). 1HNMR (600 MHz, CDCl3)
δ 7.38–7.31 (m, 4H), 7.29–7.25 (m, 3H), 7.25–7.20
(m, 3H), 7.18 (d, J = 6.0 Hz, 2H), 6.98 (d, J = 8.5 Hz, 2H), 6.93–6.88 (m, 2H), 5.95 (dtd, J = 17.2, 9.5, 4.0 Hz, 1H), 5.25 (d, J =
10.1 Hz, 1H), 5.20 (d, J = 17.2 Hz, 1H), 5.12–5.05
(m, 1H), 4.88–4.82 (m, 1H), 4.61–4.52 (m, 2H), 4.52–4.45
(m, 2H), 4.08 (dd, J = 7.4, 5.3 Hz, 1H), 3.87 (d, J = 11.7 Hz, 1H), 3.85 (s, 3H), 3.82 (d, J = 13.3 Hz, 1H), 3.66–3.61 (m, 1H), 2.96 (d, J = 13.2 Hz, 1H), 2.56 (dd, J = 13.3, 9.1 Hz, 1H). 13C{1H} NMR (151 MHz, CDCl3) δ 159.0, 158.7,
140.1, 137.1, 134.7, 131.8, 131.0, 130.0, 128.7, 128.6, 128.0, 127.8,
127.6, 119.2, 118.0, 113.8, 70.4, 62.4, 56.9, 56.7, 55.4, 52.9, 52.5.
HRMS (DART) m/z calcd for C30H33N2O3 [M + H]+: 469.2491; Found [M + H]+: 469.2504.
Synthesis
of 27
To a 20 mL crimp cap vial
with a stir bar in an Ar filled glovebox were added 48 mg (0.10 mmol)
of 26 followed by 2 mL of toluene and 3.2 mg (5.1 μmol)
of a Hoveyda–Grubbs II catalyst. The vial was sealed and removed
from the glovebox. The solution was heated at 90 °C for 12 h.
The reaction mixture was concentrated, and the crude residue was purified
by flash chromatography (50% EtOAc/hexanes) to afford 35 mg (78%)
of 27 as a colorless foam. 1HNMR (600 MHz,
CDCl3) δ 7.44 (d, J = 7.3 Hz, 2H),
7.39 (t, J = 7.5 Hz, 2H), 7.35–7.31 (m, 1H),
7.28–7.21 (m, 6H), 7.15 (d, J = 6.0 Hz, 2H),
6.88 (d, J = 7.9 Hz, 2H), 5.41–5.37 (m, 1H),
5.20 (dd, J = 8.7, 2.8 Hz, 1H), 5.15–5.09
(m, 1H), 4.72–4.68 (m, 1H), 4.34 (t, J = 6.0
Hz, 1H), 3.99 (d, J = 4.4 Hz, 1H), 3.97 (s, 1H),
3.96–3.93 (m, 1H), 3.80 (s, 3H), 3.24 (dt, J = 17.9, 2.5 Hz, 1H), 2.95 (d, J = 13.2 Hz, 1H),
2.76 (dd, J = 18.0, 2.8 Hz, 1H). 13C{1H}
NMR (151 MHz, CDCl3) δ 158.7, 157.9, 142.8, 138.2,
130.7, 129.5, 128.8, 128.7, 128.6, 128.5, 128.1, 128.0, 126.4, 123.2,
113.8, 70.7, 67.8, 59.4, 59.3, 55.3, 53.5, 51.6. HRMS (DART) m/z calcd for C28H29N2O3 [M + H]+: 441.2178; Found [M
+ H]+: 441.2205.
Authors: Eric R Welin; Aurapat Ngamnithiporn; Max Klatte; Guillaume Lapointe; Gerit M Pototschnig; Martina S J McDermott; Dylan Conklin; Christopher D Gilmore; Pamela M Tadross; Christopher K Haley; Kenji Negoro; Emil Glibstrup; Christian U Grünanger; Kevin M Allan; Scott C Virgil; Dennis J Slamon; Brian M Stoltz Journal: Science Date: 2018-12-20 Impact factor: 47.728
Figure 1
Scheme 1
Scheme 2
Scheme 3
Scheme 4
Scheme 5
Scheme 6
Figure 2
Figure 3
Figure 4
Figure 5
Scheme 7