Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Oct 1.
Published in final edited form as: Nat Chem. 2017 Apr 17;9(10):997–1004. doi: 10.1038/nchem.2760

Transition-metal-free chemo- and regioselective vinylation of azaallyls

Minyan Li †, Osvaldo Gutierrez †, Simon Berritt †, Ana Pascual-Escudero †, Ahmet Yeşilçimen †, Xiaodong Yang †, Javier Adrio †, Georgia Huang †, Eiko Nakamaru-Ogiso #, Marisa C Kozlowski †, Patrick J Walsh ‡,†
PMCID: PMC5650226  NIHMSID: NIHMS909939  PMID: 28937664

Abstract

Direct C(sp3)–C(sp2) bond-formation under transition-metal-free conditions offers an atom-economical, inexpensive, and environmentally benign alternative to traditional transition metal-catalyzed cross-coupling reactions. A new chemo- and regioselective coupling protocol between 3-aryl-substituted-1,1-diphenyl-2-azaallyl derivatives and vinyl bromides has been developed. This is the first transition-metal-free cross-coupling of azaallyls with vinyl bromide electrophiles and delivers allylic amines in excellent yields (up to 99%). This relatively simple and mild protocol offers a direct and practical strategy for the synthesis of high-value allylic amine building blocks that does not require the use of transition metals, special initiators, or photoredox catalysts. Radical clock experiments, EPR studies and DFT calculations point to an unprecedented substrate-dependent coupling mechanism. Furthermore, an EPR signal was observed when the N-benzyl benzophenone ketimine was subjected to silylamide base, supporting formation of radical species upon deprotonation. The unique mechanisms outlined herein could pave the way for new approaches to transition-metal-free C–C bond formations.

TOC image

Cross-coupling under transition-metal-free conditions is an attractive and economic alternative to traditional transition metal-catalyzed methods. Metal-free coupling of azaallyls has now been demonstrated with vinyl bromide electrophiles,delivering allylic amines in excellent yields. Moreover, evidence supporting dual reaction pathways triggered by ketimine anions and radicals is described.

graphic file with name nihms909939u1.jpg

Transition metal-catalyzed coupling reactions are among the most widely used, robust, and efficient methods for carbon-carbon and carbon-heteroatom bond-formations. The impact of these transformations has extended well beyond organic synthesis into biology, materials science, and engineering and was highlighted by the Nobel Prize in 2010.1,2 The development of transition-metal-free protocols would offer an inexpensive and environmentally benign alternative to accomplish C–C and C–heteroatom bond constructions.3 In this regard, we focused our attention on transition-metal-free assembly of allylic amines, which are key building blocks in the synthesis of complex natural products and are prevalent in biologically active compounds.4 To date, approaches to the synthesis of allylic amine derivatives generally rely on transition metals, including amination reactions,5,6 some nucleophilic additions of organometallic reagents to imines,7,8 and catalyzed Overman rearrangements,9 among others.10,11

Our team,12–15 and others,16–21 have been interested in the synthesis of amines via an umpolung approach involving the functionalization of 2-azaallyl anions. Recently, we reported the synthesis of both allylic amines and enamines through the transition-metal catalyzed arylation of 1,1-diphenyl-3-arylallyl-2-azaallyl anions (shown retrosynthetically in Figure 1a, path a).13 We envisioned an alternative route to the same allylic imines via vinylation of 1,1-diphenyl-3-aryl-2-azaallyl anions (Figure 1a, path b). Transition metal-catalyzed vinylation methods are less common,22–28 especially in the absence of preformed organometallic reagents (Figure 1b). We anticipated that the vinylation of azaallyl anions (Figure 1a, path b) would occur in a similar manner to the transition metal-catalyzed arylation of the same azaallyl anions. To our surprise, however, we found that the path b vinylation of 1,1-diphenyl-3-aryl-2-azaallyl anions occurred rapidly when no palladium catalyst was added.

Figure 1. Overview of allylic amines synthesis and vinylation reactions.

Figure 1

Open in a new tab

a. Retrosynthesis of allylic amines. b. Prior metal catalyzed vinylation reactions of carbanions generated in-situ. c. Transition metal-free vinylation of azaallyl-anions. d. Possible byproducts derived from product deprotonation/isomerization and deprotonation/cyclization.

Herein, we report the first transition-metal-free C(sp3)–C(sp2) coupling of vinyl bromide electrophiles with 2-azaallyl anions and azaallyl radicals (Figure 1c). Specifically, we describe coupling of 1,1-diphenyl-3-aryl-2-azaallyl anions with vinyl bromides to afford the E vinylation products in high yields with excellent chemo- and regioselectivity. Notably, the vinylation reaction outperforms a rapid background reaction arising from the base promoted elimination of the vinyl bromide to form a terminal alkyne. In addition, the base/solvent combination efficiently promotes the metal-free vinylation but does not deprotonate the product (Figure 1d). Thus, neither product deprotonation/isomerization, which will form a more stable conjugated byproduct, nor product deprotonation/cyclization, is observed. Additional experiments, EPR studies and calculations provide a preliminary picture of the mechanistic landscape of the observed vinylations.

Results

We reported previously the arylation of 1,1-diphenyl-3-arylallyl-2-azaallyl anions via in situ deprotonation of ketimine substrates (ArCH2N=CPh2) followed by palladium-catalyzed arylation with Ar′–X to give ArAr′CHN=CPh2.12–15 Under these conditions, deprotonation of the product was not observed, as judged by the absence of isomerized product (ArAr′C=NCHPh2).19,29 Key to preventing product deprotonation was the use of hindered bases [MN(SiMe3)2 (M = Li, Na)] at room temperature. To explore the analogous vinylation reaction, we investigated the deprotonative cross-coupling of ketimine 1a and β-bromostyrene 2a using microscale high-throughput screening (0.01 mmol) of 23 electronically diverse mono- and bidentate phosphine ligands in the presence of Pd(OAc)2, NaN(SiMe3)2 and THF at room temperature for three hours. To our surprise, a considerable amount of the targeted vinylated product (3aa) was found in the control vial, which contained no added transition-metal or ligand (see Supplementary information page 10, high throughput experimentation screenings). Validation on a larger scale (0.1 mmol) provided further support for a transition-metal-free process; specifically, reactions conducted without addition of catalyst led to the vinylation product 3aa in 51% assay yield (AY) in THF as determined by 1H NMR spectroscopy (Table 1, entry 1). Increasing the concentration from 0.1 M to 0.2 M (entry 2) resulted in an increase to 65% AY. We next tested 5 solvents [CPME (cyclopentyl methyl ether), THF, 2-Me-THF, DME (dimethoxyethane) and 1,4-dioxane, entries 3–7]. Surprisingly, the reaction was complete in just 10 min in DME, giving 98% AY of 3aa (entry 6). Further optimization of the base ratio revealed that 2 equiv of LiN(SiMe3)2 and a 30 min reaction time resulted in 98% isolated yield (entry 9).

Table 1.

Optimization of the transition metal-free vinylation.a,b

graphic file with name nihms909939u2.jpg
Entry Base (equiv) Solvent Time (min) (min) Conc. 3aa (%)
1 NaN(SiMe3)2 (3) THF 180 0.1 M 51
2 NaN(SiMe3)2 (3) THF 180 0.2 M 65
3 NaN(SiMe3)2 (3) CPME 10 0.2 M 0
4 NaN(SiMe3)2 (3) THF 10 0.2 M 32
5 NaN(SiMe3)2 (3) 2-Me THF 10 0.2 M 0
6 NaN(SiMe3)2 (3) DME 10 0.2 M 98
7 NaN(SiMe3)2 (3) Dioxane 10 0.2 M 0
8 LiN(SiMe3)2 (3) DME 10 0.2 M 99
9 LiN(SiMe3)2 (2) DME 10 0.2 M 98c
10 LiN(SiMe3)2 (1.5) DME 10 0.2 M 54
11 LiN(SiMe3)2 (1) DME 30 0.2 M 21
Open in a new tab
a

Reactions conducted on a 0.1 mmol scale.

b

Assay yields determined by 1H NMR spectroscopy of the reaction mixture on a 0.1 mmol scale using an internal standard.

c

Isolated yield after chromatographic purification.

Reaction Scope

With the optimized reaction conditions (entry 9, Table 1), the scope of the N-benzyl group on the ketimine was examined (Figure 2a). Overall, some tuning of the reaction parameters, such as the base, reagent ratios, concentration and reaction times, was necessary for certain substrates. Nonetheless, vinylation of ketimine substrates bearing electron-donating (4-Me, 1b and 4-OMe 1c) and electron-withdrawing [4-F (1d), 4-Cl (1e), 3,5-di-F (1f) and 3,5-di-CF3 (1g)] N-benzyl derived groups with β-bromostyrene (2a) proceeded in excellent yields. The sterically demanding 2-tolyl substituted ketimine (1h) underwent coupling, albeit in lower yield (61%), even when performed with 3 equiv LiN(SiMe3)2 and an extended reaction time (8 h). For more acidic heterocyclic 2-pyridyl 1i (77%), 3-pyridyl 1j (71%) and 2-thiophenyl 1k (82%) ketimine substrates, 0.05 M concentration was optimal with 1.5 equivalent of LiN(SiMe3)2 to prevent deprotonation, isomerization and cyclization (Figure 1d).

Figure 2. Overview of reaction scope.

Figure 2

Open in a new tab

a, Scope of ketimines in the vinylation reaction. aUnless otherwise noted, reactions were conducted on a 0.1 mmol scale using 1 equiv of ketimine 1 and 2 equiv of 2a at 0.2 M. Isolated yields after chromatographic purification. b2 equiv of LiN(SiMe3)2, 30 min. c3 equiv of LiN(SiMe3)2, 3 h. d2 equiv of LiN(SiMe3)2, 6 h. e0.05 M. f1.2 equiv of LiN(SiMe3)2, 0.05 M, 15 min. g2 equiv of LiN(SiMe3)2, 8 h. h0.05 M, 3 h.). b. Scope of vinyl bromides in transition metal-free vinylation. aReactions conducted on a 0.1 mmol scale using 1 equiv 1a (0.2 M), 2 equiv LiN(SiMe3)2, and 2 equiv vinyl bromides at 0.2 M. Isolated yields after chromatographic purification. b2 equiv LiN(SiMe3)2 c0.1 M. d3 equiv NaN(SiMe3)2, 12 h, 0.1 M. e3 equiv NaN(SiMe3)2, 15 min, 60 °C.). c. Gram-scale sequential one-pot imine synthesis/vinylation protocol. d. Ketimine hydrolysis.

We next determined the scope in the vinyl bromide component and found it to be broad (Figure 2b). trans-β-Aryl vinyl bromides bearing neutral (4-C6H4-tBu) and electron donating (4-C6H4-OMe) groups furnished coupling products 3ab and 3ac in 92 and 98% yield, respectively. 4-Fluoro- (2d) and 4-trifluoromethyl-substituted vinyl bromides (2e) afforded the coupled products 3ad and 3ae in 96 and 70% yield. The reactions also proceeded well with substitution at the 3-position of the β-bromostyrene, such as 3-OMe (2f) and 3-CF3 (2g), giving the products in 98 and 75% yield, respectively. trans-β-2-Tolyl vinyl bromide (2h) was an excellent substrate, affording the product in 96% yield. Surprisingly, the aliphatic vinyl bromides 1-bromo-2-methylprop-1-ene (2i) and (bromomethylene)cyclohexane (2j) were successfully coupled with 1a to afford 3ai and 3aj in 54–55% yield. Monitoring the reaction with 3aj revealed that the transformation was complete in 15 min (see Supplementary Figure 1). Notably, product isomerization and/or cyclization were not detected for any of the examples in Figure 2a–b.

The vinylation is amenable to a telescoped imine synthesis/vinylation protocol on gram-scale (Figure 2c). For example, the sequential one-pot synthesis of 2-thiophenyl ketimine 1k, followed by vinylation with 2a successfully afforded product 3ka in 90% yield (1.02 g). Finally, hydrolysis of the vinylated product 3ka was performed to isolate the allylic amine 5ka in 99% yield (Figure 2d).

Mechanistic Studies

Given the unexpected and unusual coupling of ketimines with vinyl bromides in the absence of added catalyst, a synergistic computational and experimental study was undertaken to gain insight into the mechanistic possibilities. Several reasonable mechanisms can be envisioned (Figure 3a), including those involving alkyne intermediates, cycloadditions, addition/eliminations, substitutions and carbene insertions.

Figure 3. Overview of Mechanistic Study.

Figure 3

Open in a new tab

a, Possible mechanisms for vinylation of benzyl ketimines. (i.) Elimination-addition pathway. (ii.) [3+2] cycloaddition between ketimine anion and vinyl brimode, followed by ring opening. (iii.) Carbene insertion. (iv.) Addition-elimination or substitution. b. Probing the intermediacy of an alkyne. c. Probing the vinylidine insertion mechanism.

Of relevance to pathway i in Figure 3a, in the absence of ketimine the vinyl bromide undergoes rapid (one hour at room temperature) NaN(SiMe3)2 promoted background elimination to yield the corresponding alkyne. Mixing the terminal alkyne and imine 1a under the same conditions as those used in the vinylation did not result in formation of the vinylated product 3aa (ii, Figure 3b). These results indicate that the alkyne is not an intermediate en route to the vinylation product, discounting the pathway i in Figure 3a.

Addition of the silylamide base is expected to cause rapid deprotonation of the N-benzyl ketimine giving rise to an ketimine anion (Figure 3a).30,31 It is known that azaallyl anions undergo cycloadditions with styrene and, therefore, this pathway was explored with β-bromostyrene.32,33 Two isomeric [3+2] reactions are possible from this intermediate (ii, Figure 3a), with the first leading to the observed product after elimination of bromide. However, significant steric interactions arise in this pathway between the vinyl aryl group and the diphenyl moiety of the azaallyl anion. The alternate cyclization pathway gives rise to the regioisomeric product, which was not observed. Computations undertaken on the azaallyl anion found the concerted [3+2] cyclization pathway to be higher in energy (see Supplementary Information) than the step-wise substitution mechanism described below (iv, Figure 3a).

Another possibility is the α-elimination of vinyl halides in the presence of MN(SiMe3)2 (M = Na, K) to form vinylidenes.34,35 A highly reactive vinylidene could then insert into the benzylic C–H bond of the ketimine. If this mechanism is operative in our system, the hydrogen of the vinyl motif would originate from a benzylic hydrogen of the ketimine 1a. With this in mind, the deuterated ketimine 1a′ was prepared and investigated in the coupling with vinyl bromides 2i and 2a. The results (Figure 3c) indicate that vinyl hydrogens of the products are not deuterated, which excludes the α-elimination mechanism in iii, Figure 3a.

Our attention next turned to possible substitution pathways.36–38 Both stepwise and concerted (SNV)39–42 substitutions have been reported for vinyl halides, but not with carbon-derived anions. Mechanisms are case-dependent with the nature of the nucleophile, the leaving group, and the electrophile substitution pattern all playing a role.43 All attempts to calculate intermediates for the corresponding stepwise process (iv, Figure 3a) using functionals previously used to study SNV reactions (B3LYP and OPBE) led directly to dissociation back to starting materials or formation of vinylated product and bromide. DFT calculations did, however, reveal a concerted displacement with a low energetic barrier (11.7 kcal/mol) to deliver the vinylated product B1 with concomitant release of bromide (Figure 4). A strong kinetic preference (ΔΔG‡ > 7 kcal/mol) for nucleophilic attack perpendicular to the carbon-carbon double bond (SNV)39–42 at the less sterically encumbered β-position (A1-B1-TS vs A1-C1-TS) was found. Notably, both product regioisomers (B1 and C1) are nearly isoenergetic.

Figure 4.

Figure 4

Open in a new tab

M06-2X/6-31G(d)-CH2Cl2(CPCM) geometries of competing transition states for azaallyl anion SNV reaction with styryl bromide. Gibbs free energies (blue) are in kcal/mol relative to isolated reactants. Selected distances (red) are in Å.

These computational results are in accord with both the experimentally observed reactivity (20 min – 3 h at room temperature) and regioselectivity (> 20:1 dr) using bromostyrene. However, experiments with the Z-bromostyrene 2a′ at room temperature yield a near 2:1 ratio of E- and Z-products, albeit in low yield (14%) due to rapid competing elimination of Z-bromostyrene 2a′ to form alkyne (see Supplementary Figure 5). The anionic-pathway SNV shown in Figure 4 is a concerted and stereospecific process,36–38 which is inconsistent with the stereochemical scrambling observed. These results motivated us to examine alternative pathways.

We speculated that an azaallyl radical A0 (Figure 5a)44–46 could participate in the process. This azaallyl radical could form by electron transfer from the azaallyl anion (A1) to a molecule of ketimine 1a (Figure 5a).47 DFT calculations for azaallyl radical recombination with a putative vinyl radical45,48 showed very small energetic barriers (ca. 11 kcal/mol), indicating that such a process is facile. The energy difference between the competing transition states, however, indicates negligible levels of E/Z product regioselectivity, which is inconsistent with experiments (see Supplementary Information for structures).

Figure 5. Overview of possible radical mechanism.

Figure 5

Figure 5

Figure 5

Figure 5

Open in a new tab

a. Proposed formation of ketimine radical A0 and ketimine radical anion A2 via a SET process. b. Gibbs free energy (M06-2X/6-31G(d)-CH2Cl2(CPCM)) profile for the azaallyl radical additions. Energies (kcal/mol) are relative to isolated reactants. Substituent effects of electrophile on the radical and anionic addition barriers (free energies in kcal/mol at 298 K) were calculated using M06-2X/6-31G(d)-CH2Cl2(CPCM). Selected Mulliken spin densities with hydrogens summed into heavy atoms are listed in blue. Selected distances are in Å. c. Substituent effects of electrophile on the radical and anionic addition barriers (free energies in kcal/mol at 298 K) were calculated using M06-2X/6-31G(d)-CH2Cl2(CPCM). d. Substrate study of vinylation. e. Radical clock study.

Transition states could also be located for the addition of azaallyl radical A0 to trans-bromostyrene (Figure 5b), which proceeds in a stepwise manner with an overall barrier of 18.0 kcal/mol (via TS-1a) to yield radical intermediate Int-1a, and eventually to the observed regioisomer (Figure 5b). The transition state TS-2a, which leads to the unobserved regioisomer is much higher in energy (23.2 kcal/mol) due to increased steric interactions between the diaryl moiety of the azaallyl radical and β-bromostyrene.

From Int-1a, a scan for dissociation of Br• and formation of observed product P-1a was computed to be prohibitively high in energy (>35 kcal/mol; see Supplementary Information for scan). Alternatively, radical intermediate Int-1a can undergo single electron reduction to generate a negatively charged species, which facilitates the heterolytic cleavage of the C–Br bond to generate Br− and the observed product P-1a. In support of this hypothesis, all optimizations of anionic version of Int-1a led directly to dissociation of Br− and formation of P-1a. This result implies that once radical Int-1a undergoes single electron reduction, it quickly dissociates bromide leading to the observed product, consistent with the highly exergonic (–47 kcal/mol) nature of the net process. This process also converts radical anion A2 back to ketimine 1a. The product regioisomers P-1a and P-2a are nearly isoenergetic, and hence the regioselectivity derives from the kinetic preference of TS-1a over TS-2a. In contrast to the azaallyl anion pathway (Figure 5b), the pathway involving the radical intermediates Int-1a accounts for the stereochemical scrambling from Z-bromostyrene 2a′ (bond rotation more rapid than reduction/elimination).

The mechanism in Figure 5 proposes radical intermediates (A0 and A2). To probe for the presence of radicals, an electron paramagnetic resonance (EPR) study was undertaken. In these experiments, DME solutions of NaN(SiMe3)2 were added to DME solutions of ketimine 1a and the samples were allowed to react for ~5 min at rt before freezing in liquid nitrogen. The EPR spectra were acquired in DME glass at 190 K and microwave power of 1 mW. No signal was detected in samples with only base in DME or DME alone. In contrast, for the deprotonations of 1a, EPR spectra showed clear signals for the presence of radical species (see Supplementary Figure 2). The observation of an EPR signal is supportive of one or more radicals, but definitive conclusions concerning the nature of the radical species must await future investigations. Despite numerous studies of azaallyl anions in the literature, to the best of our knowledge this is the only evidence of radical formation upon deprotonation of ketimines.

Given the evidence for the presence of radicals, which we propose to be the azaallyl radical (A0) and the ketimine radical anion (A2), an effort was made to distinguish the azaallyl anion SNV and azaallyl radical additions by calculating the reaction barriers (Figure 5c). Based on the computations, similar trends were observed for both mechanisms, as listed in Figure 5d. Thus, E-β-bromostyrene was predicted to react more readily than the Z-isomer or E-β-chlorostyrene. Experimentally, E-β-bromostyrene reacted at –40 °C to afford the vinylation product in 96% assay yield (i, Figure 5d) while Z-β-bromostyrene generated the E-product in 6% yield (no cis-product observed and elimination dominating the reaction, ii, Figure 5d). At room temperature, coupling of E-β-chlorostyrene afforded product in 98% assay yield. For substrates lacking a β-aryl group, reaction is expected to slow according to both mechanisms, but to a far larger degree for the azaallyl anion mechanism.39–42 As illustrated in Figure 2b, 1-bromo-2-methylprop-1-ene (2i) and (bromomethylene)cyclohexane (2j) underwent reaction, albeit more slowly (12 h) and with reduced yields (54 and 55%, respectively). Although there are well-documented errors associated with determining accurate quantitative barriers with charged species, computations employing the commonly used B3LYP functional, which has been used by others to study SNV reactions, predicted prohibitively high barriers (>36 kcal/mol) for the azaallyl anion A1 and 1-bromo-2-methylprop-1-ene in both gas-phase and in implicit solvent. Moreover, the greater reactivity of 1-bromo-2-methylprop-1-ene (Figure 2b, 2i) vs E-1-bromo-prop-1-ene (iv, Figure 5d) also seemingly runs counter to the azaallyl anion mechanism; a build-up of negative charge would be less favorable with two methyl substituents as is the case for the former.

The EPR results support the presence of radicals, but do not indicate if they are involved in the vinylation reactions. To probe for the direct participation of radicals in the vinylations, we next examined the reactions of radical clock-containing styryl and non-styryl substrates. We prepared the radical clock 2k (Figure 5e) to test for the intermediacy of radical intermediates in the reaction with β-bromostyrene derivatives. As outlined in Figure 5b, Int-1a possesses a stabilized benzylic radical. Thus, a second aryl group was built into the substrate such that, in the event the benzylic radical analogous to Int-1a is formed, ring-opening would also give a stabilized benzylic radical. The reaction of the E-styryl radical clock 2k with ketimine 1a in the presence of LiN(SiMe3)2 furnished a mixture of Z and E vinylated products in a 1:1 ratio (i, Figure 5e, overall 60% yield). An alkyne side-product was observed in 28% isolated yield, which we propose is derived from the α-elimination-migration mechanism. Thus, no cyclopropane ring-opened products were detected, suggesting that the reaction of styryl substrates does not proceed through radical intermediates. Furthermore, the scrambling of the double bond geometry is inconsistent with an anionic concerted SNV substitution mechanism. Based on these results, the pathway favored is the addition/elimination, wherein the azaallyl anion adds to the styrene to generate a benzylic carbanion that can rotate about the single bond before elimination, giving rise to the observed products (see structure of the proposed intermediate in i, Figure 5e). Unfortunately, we were unable to locate such an intermediate computationally without an additional anion-stabilizing group.

To probe the participation of radicals with aliphatic vinyl bromides, radical clock 2l was prepared and isolated as the pure E-isomer (see Supplementary Information page 19, synthesis of radical clock 2l for details). Reaction with of E-2l (ii, Figure 5e) with ketimine 1a in the presence of LiN(SiMe3)2 led to a complex mixture, as determined by 1H NMR. Further, no clean products could be isolated from this mixture. We propose that ring-opening occurs to give a reactive radical. These results suggest the presence of radical addition mechanism for aliphatic vinyl bromides.

Based on the computational, spectroscopic (EPR), and experimental results, both azaallyl anion and radical exist under the reaction conditions. To rationalize these observations, we propose substrate dependent reactions with different electrophiles. Thus, with aliphatic vinyl bromides the azaallyl radical addition mechanism best fits with the experimental and computational results. In contrast, for the styryl vinyl bromides, the data is most consistent with the anionic pathway.

Conclusion

In summary, we report a unique transition-metal free coupling of 1,1-diphenyl-3-aryl-2-azaallyl anions with vinyl bromides to afford allylic amine derivatives, which are of great importance in the pharmaceutical industry. High regioselectivities and yields of E-vinylation products were observed. The vinylation reaction outcompetes a rapid background reaction arising from the elimination of the styryl bromides to form alkynes. Several mechanistic possibilities were explored both experimentally and with DFT calculations. EPR studies indicate the presence of radicals, which we have tentatively assigned to be the azaallyl radical and imino-ketyl radical anion species. The detection of radicals is consistent with the unusual azaallyl radical addition mechanism with aliphatic vinyl bromides. For styryl bromides, a radical clock-bearing substrate reacted without ring opening, but with loss of the double bond stereochemistry. This observation disfavors the radical mechanism and is better explained by an anionic addition-elimination mechanism. Further studies are underway to broaden this class of reactions utilizing ketimine derivatives in the absence of added transition metal catalyst, and definitively assign the structures of the radical species observed by EPR.

Supplementary Material

Supporting Information

Acknowledgments

We thank the National Science Foundation (CHE-1464744 to P. J. W. and CHE-1464778 to M. C. K.) and National Institutes of Health (GM-104349 to P. J. W. and GM-087605 to M. C. K.) for financial support. Computational support was provided by XSEDE on SDSC Gordon (TG-CHEM120052). This work was also financially supported by SICAM Fellowship by Jiangsu National Synergetic Innovation Center for Advanced Materials. We thank Dr. Sonia Montel of UPenn and Prof. Karl Scheidt of Northwestern University for helpful discussions.

Footnotes

Data availability statement

All relevant data are included with the manuscript, supplementary information and are available from the corresponding authors on reasonable request.

Author contributions

M.L. and P.J.W. conceived and designed the experiments. M.L., S.B., A.P.E, A.Y., X.Y., J.A., G.H. performed the research. O.G. and M.C.K. designed and performed the DFT computational study. M.L. and E.N.O. performed the EPR study. M.L., O.G., M.C.K. and P.J.W. wrote the paper.

References

  • 1.Negishi E. Magical power of transition metals: past, present, and future (Nobel Lecture) Angew Chem Int Ed. 2011;50:6738–6764. doi: 10.1002/anie.201101380. [DOI] [PubMed] [Google Scholar]
  • 2.Suzuki A. Cross-coupling reactions of organoboranes: an easy way to construct C-C bonds (Nobel Lecture) Angew Chem Int Ed. 2011;50:6722–6737. doi: 10.1002/anie.201101379. [DOI] [PubMed] [Google Scholar]
  • 3.Sun CL, Shi ZJ. Transition-metal-free coupling reactions. Chem Rev. 2014;114:9219–9280. doi: 10.1021/cr400274j. [DOI] [PubMed] [Google Scholar]
  • 4.Petranyi G, Ryder N, Stutz A. Allylamine derivatives: new class of synthetic antifungal agents inhibiting fungal squalene epoxidase. Science. 1984;224:1239–1241. doi: 10.1126/science.6547247. [DOI] [PubMed] [Google Scholar]
  • 5.Roggen M, Carreira EM. Stereospecific substitution of allylic alcohols to give optically active primary allylic amines: unique reactivity of a (P,alkene)Ir complex modulated by iodide. J Am Chem Soc. 2010;132:11917–11919. doi: 10.1021/ja105271z. [DOI] [PubMed] [Google Scholar]
  • 6.Yamashita Y, Gopalarathnam A, Hartwig JF. Iridium-catalyzed, asymmetric amination of allylic alcohols activated by Lewis acids. J Am Chem Soc. 2007;129:7508–7509. doi: 10.1021/ja0730718. [DOI] [PubMed] [Google Scholar]
  • 7.Patel SJ, Jamison TF. Asymmetric catalytic coupling of organoboranes, alkynes, and imines with a removable (trialkylsilyloxy)ethyl group–direct access to enantiomerically pure primary allylic amines. Angew Chem Int Ed. 2004;43:3941–3944. doi: 10.1002/anie.200460044. [DOI] [PubMed] [Google Scholar]
  • 8.Shi X, Kiesman WF, Levina A, Xin Z. Catalytic asymmetric petasis reactions of vinylboronates. The Journal of organic chemistry. 2013;78:9415–9423. doi: 10.1021/jo4016425. [DOI] [PubMed] [Google Scholar]
  • 9.Overman LE. A general method for the synthesis of amines by the rearrangement of allylic trichloroacetimidates. 1,3 Transposition of alcohol and amine functions. J Am Chem Soc. 1976;98:2901–2910. [Google Scholar]
  • 10.Ngai MY, Barchuk A, Krische MJ. Enantioselective Iridium-Catalyzed Imine Vinylation:  Optically Enriched Allylic Amines via Alkyne−Imine Reductive Coupling Mediated by Hydrogen. J Am Chem Soc. 2007;129:12644–12645. doi: 10.1021/ja075438e. [DOI] [PubMed] [Google Scholar]
  • 11.Skucas E, Kong JR, Krische MJ. Enantioselective reductive coupling of acetylene to N-arylsulfonyl imines via Rhodium catalyzed C−C bond-forming hydrogenation:  (Z)-dienyl allylic amines. J Am Chem Soc. 2007;129:7242–7243. doi: 10.1021/ja0715896. [DOI] [PubMed] [Google Scholar]
  • 12.Li M, Berritt S, Walsh PJ. Palladium-catalyzed regioselective arylation of 1,1,3-triaryl-2-azaallyl anions with aryl chlorides. Org Lett. 2014;16:4312–4315. doi: 10.1021/ol502043j. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Li M, et al. Palladium-catalyzed C−H arylation of α,β-unsaturated imines: catalyst-controlled synthesis of enamine and allylic amine derivatives. Angew Chem Int Ed. 2016;128:2875–2879. doi: 10.1002/anie.201509757. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Li M, Yucel B, Adrio J, Bellomo A, Walsh PJ. Synthesis of diarylmethylamines palladium-catalyzed regioselective arylation of 1,1,3-triaryl-2-azaallyl anions. Chem Sci. 2014;5:2383–2391. doi: 10.1039/C3SC53526F. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Li M, et al. Umpolung synthesis of diarylmethylamines via Palladium-catalyzed arylation of N-benzyl aldimines. Adv Synth Catal. 2016;358:1910–1915. doi: 10.1002/adsc.201600075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Wu Y, Hu L, Li Z, Deng L. Catalytic asymmetric umpolung reactions of imines. Nature. 2015;523:445–450. doi: 10.1038/nature14617. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Burger EC, Tunge JA. Synthesis of homoallylic amines via the Palladium-catalyzed decarboxylative coupling of amino acid derivatives. J Am Chem Soc. 2006;128:10002–10003. doi: 10.1021/ja063115x. [DOI] [PubMed] [Google Scholar]
  • 18.Fields WH, Chruma JJ. Palladium-catalyzed decarboxylative benzylation of diphenylglycinate Imines. Organic letters. 2009;12:316–319. doi: 10.1021/ol902651j. [DOI] [PubMed] [Google Scholar]
  • 19.Niwa T, Yorimitsu H, Oshima K. Palladium-catalyzed benzylic arylation of N-benzylxanthone imine. Organic letters. 2008;10:4689–4691. doi: 10.1021/ol802070d. [DOI] [PubMed] [Google Scholar]
  • 20.Zhu Y, Buchwald SL. Ligand-controlled asymmetric arylation of aliphatic α-amino anion equivalents. J Am Chem Soc. 2014;136:4500–4503. doi: 10.1021/ja501560x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Liu YE, et al. Enzyme-inspired axially chiral pyridoxamines armed with a cooperative lateral amine chain for enantioselective biomimetic transamination. J Am Chem Soc. 2016;138:10730–10733. doi: 10.1021/jacs.6b03930. [DOI] [PubMed] [Google Scholar]
  • 22.Ankner T, Cosner CC, Helquist P. Palladium- and Nickel-catalyzed alkenylation of enolates. Chem Eur J. 2013;19:1858–1871. doi: 10.1002/chem.201202798. [DOI] [PubMed] [Google Scholar]
  • 23.Grigalunas M, Ankner T, Norrby PO, Wiest O, Helquist P. Palladium-catalyzed alkenylation of ketone enolates under mild conditions. Organic letters. 2014;16:3970–3973. doi: 10.1021/ol5017965. [DOI] [PubMed] [Google Scholar]
  • 24.Hardegger LA, Habegger J, Donohoe TJ. Modular synthesis of highly substituted pyridines via enolate alpha-alkenylation. Org Lett. 2015;17:3222–3225. doi: 10.1021/acs.orglett.5b01312. [DOI] [PubMed] [Google Scholar]
  • 25.Padilla-Salinas R, Walvoord RR, Tcyrulnikov S, Kozlowski MC. Nitroethylation of vinyl triflates and bromides. Org Lett. 2013;15:3966–3969. doi: 10.1021/ol401747u. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Piers E, Marais PC. A new five-membered ring annulation method based on Palladium(0)-catalyzed intramolecular coupling of vinyl iodide and enolate anion functions. J Org Chem. 1990;55:3454–3455. [Google Scholar]
  • 27.Solé D, Peidró E, Bonjoch J. Palladium-catalyzed intramolecular coupling of vinyl halides and ketone enolates. synthesis of bridged azabicyclic compounds. Org Lett. 2000;2:2225–2228. doi: 10.1021/ol005973i. [DOI] [PubMed] [Google Scholar]
  • 28.Yang X, Kim BS, Li M, Walsh PJ. Palladium-catalyzed selective alpha-alkenylation of pyridylmethyl ethers with vinyl bromides. Org Lett. 2016;18:2371–2374. doi: 10.1021/acs.orglett.6b00815. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Niwa T, Suehiro T, Yorimitsu H, Oshima K. Carbon–carbon bond formations at the benzylic positions of N-benzylxanthone imines and N-benzyldi-1-naphthyl ketone imine. Tetrahedron. 2009;65:5125–5131. [Google Scholar]
  • 30.Tang S, Park JY, Yeagley AA, Sabat M, Chruma JJ. Decarboxylative generation of 2-azaallyl anions: 2-iminoalcohols via a decarboxylative erlenmeyer reaction. Org Lett. 2015;17:2042–2045. doi: 10.1021/acs.orglett.5b00107. [DOI] [PubMed] [Google Scholar]
  • 31.Yeagley AA, Lowder MA, Chruma JJ. Tandem C−C bond-forming processes: interception of the Pd-catalyzed decarboxylative allylation of allyl diphenylglycinate imines with activated olefins. Org Lett. 2009;11:4022–4025. doi: 10.1021/ol901745x. [DOI] [PubMed] [Google Scholar]
  • 32.Kauffmann T, Berger D, Scheerer B, Woltermann A. Anionic 3+ 2 Cycloaddition of a 1,2-Diazaallyllithium Compound. Angew Chem Int Ed. 1970;9:961–962. [Google Scholar]
  • 33.Pandiancherri S, Lupton DW. Preparation of 2-azaallyl anions and imines from N-chloroamines and their cycloaddition and allylation. Tetrahedron Lett. 2011;52:671–674. [Google Scholar]
  • 34.Taber DF, Sahli A, Yu H, Meagley RP. Efficient intramolecular C-H insertion by an alkylidene carbene generated from a vinyl chloride. J Org Chem. 1995;60:6571–6573. [Google Scholar]
  • 35.Taber DF, Sikkander MI, Storck PH. Enantioselective synthesis of (+)-majusculone. J Org Chem. 2007;72:4098–4101. doi: 10.1021/jo070257g. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Bernasconi CF, Rappoport Z. Recent advances in our mechanistic understanding of S(N)V reactions. Acc Chem Res. 2009;42:993–1003. doi: 10.1021/ar900048q. [DOI] [PubMed] [Google Scholar]
  • 37.Rappoport Z. Nucleophilic vinylic substitution. A single- or a multi-step process? Acc Chem Res. 1981;14:7–15. [Google Scholar]
  • 38.Rappoport Z. The rapid steps in nucleophilic vinylic addition-elimination substitution. Recent developments. Acc Chem Res. 1992;25:474–479. [Google Scholar]
  • 39.Bach RD, Baboul AG, Schlegel HB. Inversion versus retention of configuration for nucleophilic substitution at vinylic carbon. J Am Chem Soc. 2001;123:5787–5793. doi: 10.1021/ja010234y. [DOI] [PubMed] [Google Scholar]
  • 40.Castro EA, Gazitua M, Santos JG. Kinetics and mechanism of the anilinolysis of aryl 4-nitrophenyl carbonates in aqueous ethanol. J Org Chem. 2005;70:8088–8092. doi: 10.1021/jo051168b. [DOI] [PubMed] [Google Scholar]
  • 41.Castro EA, Ramos M, Santos JG. Concerted pyridinolysis of aryl 2,4,6-trinitrophenyl carbonates. J Org Chem. 2009;74:6374–6377. doi: 10.1021/jo901137f. [DOI] [PubMed] [Google Scholar]
  • 42.Williams A. Concerted mechanisms of acyl group transfer reactions in solution. Acc Chem Res. 1989;22:387–392. [Google Scholar]
  • 43.Fernandez I, Bickelhaupt FM, Uggerud E. Reactivity in nucleophilic vinylic substitution (SNV):SNVπ versus SNVσ mechanistic dichotomy. J Org Chem. 2013;78:8574–8584. doi: 10.1021/jo401242f. [DOI] [PubMed] [Google Scholar]
  • 44.Baum AA, Karnischky LA. Photochemical formation of oxazolidines from aryl ketones and aliphatic imines. J Am Chem Soc. 1973;95:3072–3074. [Google Scholar]
  • 45.Dannenberg JJ, Tanaka K. Theoretical studies of radical recombination reactions. 1 Allyl and azaallyl radicals. J Am Chem Soc. 1985;107:671–674. [Google Scholar]
  • 46.Malassa A, Agthe C, Görls H, Friedrich M, Westerhausen M. Deprotonation and dehydrogenation of di(2-pyridylmethyl)amine with M[N(SiMe3)2]2 (M = Mn, Fe, Co, Zn) and Fe(C6H2-2,4,6-Me3)2. J Organomet Chem. 2010;695:1641–1650. [Google Scholar]
  • 47.Pallagi I, Toró A, Horváth G. Mechanism of the gibbs reaction. part 4.1indophenol formation via N-chlorobenzoquinone imine radical anions The aza-SRN2 chain reaction mechanism chain initiation with 1,4-benzoquinones and cyanide ion. J Org Chem. 1999;64:6530–6540. doi: 10.1021/jo982113v. [DOI] [PubMed] [Google Scholar]
  • 48.Giese B. The Stereoselectivity of Intermolecular Free Radical Reactions [New Synthetic Methods (78)] Angew Chem Int Ed. 1989;28:969–980. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information
NIHMS909939-supplement-Supporting_Information.pdf (4.3MB, pdf)

RESOURCES