Introduction

Allene moieties have been found in many natural products and pharmaceuticals1,2,3,4. Due to the unique reactivities, allenes have also been demonstrated as useful and important starting materials in organic synthesis5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25. Thus, efficient methods for the synthesis of allenes from simple and readily available chemicals are highly desirable26,27,28,29,30,31,32,33,34,35,36. For such a reaction, the most straight forward method is using the allenylation of terminal alkynes (ATA) reaction due to the fact that all the starting materials, that is, terminal alkynes, amines, carbonyl compounds (aldehydes and ketones), are common chemicals in any chemical laboratory37,38,39,40,41. However, so far this ATA reaction may only be applied to paraformaldehyde (with CuI)37,38,39,40 and aldehydes (with ZnII or Cu(I))41 for the synthesis of mono- and 1,3-disubstituted allenes. The synthesis of trisubstituted allenes from this approach using ketones is still not possible (Fig. 1)42,43.

Figure 1: Approaches to allenes from terminal alkynes.
figure 1

(a) Allenylation using formaldehdye. (b) Allenylation using aldehydes. (c) Synthesis of trisubstituted allenes using ketones.

In this paper, we report a method for the synthesis of trisubstituted allenes using a CdI2 (refs 44, 45)-mediated ATA reaction of ketones (Fig. 1). With the use of CdI2, a series of trisubstituted allenes can be prepared easily in a one-pot synthesis. By applying this protocol, 1,5-bisallenes and optically active allenols, which are useful in organic synthesis, can also be prepared successfully.

Results

Different inorganic metallic salts for the ATA reaction

Our initial work began with commercially available starting materials: phenylacetylene 1a, 2-hexanone 2a and pyrrolidine 3a under the mediation of different groups 11 and 12 inorganic metallic salts (Table 1). After systematic trials, we were happy to find that the ATA reaction of 1a (1.0 mmol) and 2a (1.1 mmol) in the presence of pyrrolidine 3a (1.1 mmol) with CdI2 (0.8 mmol) in 5 ml of toluene with stirring at 130 °C for 4 h afforded the trisubstituted allene 4a in 57% yield (Table 1, entry 4)! Other metallic salts in groups 11 and 12, such as CuI (refs 37, 38, 39, 40, 46), ZnI2 (ref. 41), AgI (ref. 47), AuI (ref. 48) and HgCl2, afforded only a trace amount of the expected trisubstituted allene 4a (Table 1, entries 1–3, 5 and 6).

Table 1 Group 11 or 12 metal salts for ATA reaction*.

Optimization of the ATA reaction

The reaction was sensitive to the concentration of the substrates (Table 2, entries 1–3). With 0.6 equiv. of CdI2, the yield dropped dramatically (Table 2, entry 4). We also examined the solvent effect, with toluene still being the best (Table 2, entry 5). The yield of trisubstituted allene 4a with CdBr2 was only 7% (Table 2, entry 6). Thus, 1a (1.0 mmol), 2a (1.1 mmol), 3a (1.1 mmol) and CdI2 (0.8 mmol) in 5 ml of toluene with stirring at 130 °C for 4 h were defined as the optimized reaction conditions for further study, as the same reaction at 120 °C gave a lower yield (46%).

Table 2 Optimization of ATA reaction conditions*.

Substrate scope

With the optimal reaction conditions in hand, we then investigated the scope of the reaction by applying various terminal alkynes and ketones (Table 3). First, differently substituted terminal aryl-substituted alkynes were examined. Phenylacetylene 1a and the analogues substituted with p-MeO, p-Br, m-Br and o-Cl groups all afforded good yields of the products, providing further opportunity of elaboration (Table 3, entries 1–5). Heteroaryl-substituted acetylene 1f was also suitable for this reaction with 41% yield (Table 3, entry 6). The reaction may also be extended to terminal alkyl-substituted alkynes such as 1-octyne 1g and 1-decyne 1h, which gave trisubstituted allenes 4h and 4i in 82% and 80% yields, respectively (Table 3, entries 7 and 8). Next, we turned to investigate the ATA reaction with different ketones. 2-Hexanone 2a reacted with phenylacetylene 1a to give the corresponding allene 4a in 55% yield (Table 3, entry 9); 2-octanone 2c and 4-phenyl-2-butanone 2d worked well with 1-octyne 1g and 1-decyne 1h to afford the corresponding allenes 4k and 4l in 63% and 54% yields, respectively (Table 3, entries 11 and 12).

Table 3 The scope of terminal alkynes and ketones*.

Synthesis of 1,5-bisallenes

As we know, 1,5-bisallenes are useful starting materials in organic synthesis20,49. With this methodology, 1,5-bisallenes 4m and 4n could be easily prepared in 69% and 72% yields (Fig. 2).

Figure 2: Synthesis of 1,5-bisallenes with ATA reaction.
figure 2

1,5-Bisallenes were synthesized by applying this protocol, which are not easily prepared by other methods. NTs=N-tosyl.

Synthesis of acetyl-protected α- and β-allenols

Recently, allenols have been proven as versatile building blocks in organic synthesis5,50,51,52,53. By applying this protocol, acetate-protected trisubstituted α- and β-allenols could be obtained in moderate to good yields (Fig. 3).

Figure 3: Synthesis of acetyl-protected trisubstituted α- and β-allenols.
figure 3

Easily available acetyl-protected propargyl alcohols could be converted to trisubstituted allenes using ATA reaction.

Preparation of trisubstituted allene on gram scale

It is easy to conduct the reaction on a 1-g scale to afford 4r in 50% yield, which could be easily deprotected by its treatment with K2CO3 in MeOH and H2O to afford the corresponding allenol 4s in 96% yield, indicating the potential synthetic utility of this method (Fig. 4).

Figure 4: Gram-scale synthesis of 5,5-disubstituted allenol 4r.
figure 4

The reaction could be conducted on a 1-g scale to afford acetyl-protected trisubstituted allene, which could be converted to allenol easily.

Trisubstituted allenes with central and/or axial chirality

When we used the tert-butyldimethylsilyl ether of chiral alkynol (S)-1m (>99% ee) as the starting material, optically active tert-butyldimethylsilyl-protected α-allenol (S)-4t was also afforded in 87% yield with 99% ee (Fig. 5a). Nonsymmetrical ketone 2a was also suitable for this reaction yielded a pair of diastereoisomers (d.r.=1:1) in 69%, indicating that there is no 1,2-chiral induction of the propargylic central chirality on the in situ formed axial chirality (Fig. 5b).

Figure 5: ATA reaction with protected terminal propargylic alcohols.
figure 5

(a) When TBS-protected chiral alkynol was used, the chirality was kept in the allene product. (b) When nonsymmetrical keone 2a reacted with alkynol (±)−1 m, a pair of diastereoisomers were observed (d.r.=1:1). TBS, tert-butyldimethylsilyl.

Intermediates

It is observed that the reaction of phenylacetylene 1a, 2-octanone 2c and pyrrolidine 3a at a lower temperature, 100 °C, for 1 h afforded the corresponding propargylic amine 5j in 24% isolated yield with 3% 1H NMR yield of allene 4j. Then propargylic amine 5j was converted to allene 4j in 82% yield under the standard reaction conditions, indicating the intermediacy of 5j-type propargylic amine in this type of reaction. What is more, the propargylic amines 5u–5x may be converted to the corresponding allenes 4u–4x in 45–88% yields in the presence of CdI2 in toluene at 130 °C within 4 h (Fig. 6).

Figure 6: Synthesis of allenes from propargylic amines.
figure 6

(a) The intermediate, propargylic amine 5j, was synthesized at 100 °C and then converted to the allene successfully. (b) Allenylation of propargylic amines.

Discussion

From the results in Table 1, it should be noted that only in the case of CuI, propargylic amine 5a was formed in 73% yield, which could not be converted to allene 4a under the mediation of CuI (Table 1, entry 1), demonstrating that only CdI2 is working for both the propargylic amine and allene formation steps. In fact, with this notion that CuI could mediate the formation of propargylic amine 5a, which was considered as the intermediate, a number of metallic salts, such as ZnI2, AgI, AuI and HgCl2, were tried in this reaction in the presence of CuI (Table 4). Unfortunately, none of these mixed systems could mediate the formation of allene 4a in decent yield, indicating the unique effect of CdI2 on this ATA reaction.

Table 4 Mixed metal salts for ATA reaction*.

Based on the above data and recent reports of this type of reactions41,47, we proposed a plausible mechanism for the ATA reaction (Fig. 7): Alkynyl cadmium species 6, generated from terminal alkyne in the presence of pyrrolidine, would react with the ketoniminum 7 formed in situ from ketone and pyrrolidine, to give the corresponding propargylic amine 5. The carbon−carbon triple bond in propargylic amine 5 coordinates to CdI2 forming complex 8, which would be followed by 1,5-hydride transfer and β-elimination to afford the corresponding trisubstituted allene 4.

Figure 7: Proposed mechanism for the ATA reaction of ketones.
figure 7

The key step of propargylic amine formation followed by 1,5-hydride transfer and β-elimination. The intermediate, propargylic amine 5j, was synthesized at a lower temperature and could be converted to the allene successfully.

In conclusion, we have developed an efficient CdI2-mediated ATA reaction to synthesize trisubstituted allenes from commercially available alkynes and ketones. This protocol is operationally simple and all the starting materials are basic chemicals readily available, showing the potential synthetic utility of this method. Further studies including the asymmetric version of this reaction and the application of these allenes are being conducted in our laboratory.

Methods

Materials

Chloroform-d (Cambridge Isotope Ltd.) for 1H and 13C NMR spectroscopy was used as received. CdI2 (99.5%) was purchased from Aladdin and kept in a glove box. Toluene was dried over sodium wire with benzophenone as indicator and distilled freshly before use. Other reagents were used without further treatment.

General spectroscopic methods

1H NMR spectra were obtained at 27 °C using an Agilent spectrometer operating at 400 MHz. 13C NMR spectra were obtained at 27 °C using an Aglient spectrometer operating at 100 MHz. 1H NMR, 13C NMR and HPLC spectra are supplied for all compounds: see Supplementary Figures S1–S56. See Supplementary Methods for the characterization data of compounds not listed in this part.

Synthesis of 1-phenyl-3-methyl-1,2-heptadiene 4a

To a dried Schlenk tube was added CdI2 (293.4 mg, 0.8 mmol) inside a glove box. The Schlenk tube was then taken out and dried under vacuum with a heating gun until the white CdI2 turned to yellow green. 1a (101.5 mg, 1 mmol)/toluene (2.5 ml), 2a (109.8 mg, 1.1 mmol)/toluene (2.5 ml) and 3a (92.0 μl, d=0.852 g ml−1, 78.3 mg, 1.1 mmol) were then added sequentially under Ar atmosphere. The Schlenk tube was then equipped with a condenser and placed in a pre-heated oil bath of 130 °C with stirring for 4 h. After cooling to room temperature, the crude reaction mixture was filtrated through a short pad of silica gel eluted with ether (20 ml). After evaporation, the residue was purified by chromatography on silica gel to afford 4a (ref. 54) (102.7 mg, 55%) (eluent: petroleum ether) as a liquid: thin-layer chromatography (TLC) (petroleum ether): Rf=0.92; 1H NMR (400 MHz, CDCl3) δ 7.31−7.23 (m, 4H, Ar−H), 7.19−7.11 (m, 1H, Ar−H), 6.07−6.01 (m, 1H, CH=C), 2.12−2.02 (m, 2H, CH2), 1.79 (d, J=2.8 Hz, 3H, CH3), 1.52−1.29 (m, 4H, 2 × CH2), 0.89 (t, J=7.2 Hz, 3H, CH3); 13C NMR (100 MHz, CDCl3) δ 202.6, 136.1, 128.4, 126.5, 126.3, 103.7, 93.7, 33.8, 29.7, 22.4, 18.8, 13.9; MS (EI) m/z (%) 186 (M+, 1.61), 129 (100); IR (neat): ν=2957, 2927, 2858, 1951, 1598, 1496, 1462, 1259, 1071, 1025, cm−1.

Gram-scale synthesis of acetyl-protected β-allenol 4r

To a dried three-necked bottle was added CdI2 (2.9307, g, 8 mmol) inside a glove box. The three-necked bottle was then taken out and dried under vacuum with a heating gun until the white CdI2 turned to yellow-green. 1l (1.1250, g, 10 mmol)/toluene (20 ml), 2c (1.4100, g, 11 mmol)/toluene (20 ml) and 3a (0.7839, g, 11 mmol)/toluene (10 ml) were then added sequentially under Ar atmosphere. The three-necked bottle was then equipped with a condenser and placed in a pre-heated oil bath of 130 °C with stirring. After 5 h, the reaction was complete as monitored by TLC, the resulting mixture was cooled to room temperature and filtrated through a short pad of silica gel eluted with ether (100 ml). After evaporation, the residue was purified by chromatography on silica gel to afford 4r (1.1283, g, 50%) (eluent: petroleum ether/ethyl ether=30/1) as a liquid: TLC (petroleum ether:ethyl acetate, 10:1 v/v): Rf=0.80; 1H NMR (400 MHz, CDCl3) δ 5.03−4.94 (m, 1H, CH=C=C), 4.10 (t, J=6.8 Hz, 2H, OCH2), 2.28 (q, J=6.7Hz, 2H, CH2), 2.04 (s, 3H,CH3), 1.91 (td, J=7.4 Hz, 2.5 Hz, 2H, CH2), 1.66 (d, J=2.4 Hz, 3H, CH3), 1.44−1.22 (m, 8H, 4 × CH2), 0.88 (t, J=7.0 Hz, 3H, CH3); 13C NMR (100 MHz, CDCl3) δ 201.9, 171.0, 100.3, 85.6, 63.8, 33.8, 31.7, 28.9, 28.6, 27.4, 22.6, 20.9, 19.1, 14.0; MS (EI) m/z (%) 224 (M+, 1.01), 94 (100); IR (neat): ν=2956, 2926, 2856, 1966, 1742, 1459, 1383, 1364, 1231, 1035, cm−1; HRMS (m/z): [M]+ calcd. for C14H24O2, 224.1776; found, 224.1779.

Synthesis of optically active allene (S)-4t from chiral alkyne

Following the synthesis procedure of 4a: The reaction of CdI2 (293.7 mg, 0.8 mmol), (S)-1m (246.6 mg, 1 mmol)/toluene (2.5 ml), 2b (108.6 mg, 1.1 mmol)/toluene (2.5 ml) and 3a (92.0 μl, d=0.852 g/ml, 78.3 mg, 1.1 mmol) afforded (S)-4t (285.1 mg, 87%) (eluent: petroleum ether) as a liquid: TLC (petroleum ether): Rf=0.93; 99% ee (HPLC conditions: Chiralcel OJ−H column, CO2/i-PrOH=98/2, 1.5 ml/min, λ=214 nm, tR(major)=7.2 min, tR(minor)=6.9 min); [α]D21.7= +132.7 (c=1.01, CHCl3); 1H NMR (400 MHz, CDCl3) δ 7.36 (d, J=7.6 Hz, 2H, Ar−H), 7.30 (t, J=7.6 Hz, 2H, Ar−H), 7.21 (t, J=7.2 Hz, 1H, Ar−H), 5.21 (d, J=7.2 Hz, 1H, OCH), 5.12-5.04 (m, 1H, CH=C=C), 2.16−2.07 (m, 4H, 2 × CH2), 1.63−1.46 (m, 6H, 3 × CH2), 0.93 (s, 9H, 3 × CH3), 0.11 (s, 3H, CH3), 0.07 (s, 3H, CH3); 13C NMR (100 MHz, CDCl3) δ 197.6, 144.8, 127.9, 126.7, 125.7, 104.1, 94.7, 74.6, 31.4, 31.1, 27.3, 27.2, 26.0, 25.9, 18.3, −4.4, −4.9; MS (EI) m/z (%) 328 (M+, 1.01), 221 (100); IR (neat): ν=2928, 2887, 2855, 1966, 1492, 1469, 1447, 1252, 1084, 1062, 1006, cm−1; HRMS (m/z): [M]+ calcd. for C21H32OSi, 328.2222; found, 328.2225.

Synthesis of the intermediate 5j

To a dried Schlenk tube was added CdI2 (293.8 mg, 0.8 mmol) inside a glove box. The Schlenk tube was then taken out and dried under vacuum with a heating gun until the white CdI2 turned to yellow green. 1a (102.6 mg, 1 mmol)/toluene (2.5 ml), 2c (141.3 mg, 1.1 mmol)/toluene (2.5 ml) and 3a (92.0 μl, d=0.852 g ml−1, 78.3 mg, 1.1 mmol) were then added sequentially under Ar atmosphere. The Schlenk tube was then equipped with a condenser and placed in a pre-heated oil bath of 100 °C with stirring for 1 h. After cooling to room temperature, the crude reaction mixture was filtrated through a short pad of silica gel eluted with acetone (20 ml). After evaporation, to the residue was added 27.0 μl of CH3NO2 as the internal standard for 1H NMR analysis (29% 1H NMR yield of 5j and 3% 1H NMR yield of 4j) and then purified by chromatography on silica gel to afford 5j (67.5 mg, 24%) (eluent:petroleum ether/ethyl acetate=200:2 ml to 200:4 ml, then 20:1, finally petroleum ether/ethyl acetate/Et3N=500 ml/50 ml/0.5 ml was applied to get the pure product) as a liquid:TLC (petroleum ether:ethyl acetate, 10:1 v/v): Rf=0.20; 1H NMR (400 MHz, CDCl3) δ 7.45−7.38 (m, 2H, Ar−H), 7.33−7.23 (m, 3H, Ar−H), 2.85−2.73 (m, 4H, CH2NCH2), 1.83−1.72 (m, 5H, 2 × CH2+one proton of CH2), 1.71−1.61 (m, 1H, one proton of CH2), 1.59−1.41 (m, 5H, CH2+CH3), 1.37−1.28 (m, 6H, 3 × CH2), 0.89 (t, J=6.4 Hz, 3H, CH3); 13C NMR (100 MHz, CDCl3) δ 131.7, 128.1, 127.6, 123.6, 91.3, 84.3, 57.9, 47.7, 41.5, 31.8, 29.7, 25.8, 24.3, 23.6, 22.6, 14.1; MS (ESI) m/z 284 (M+H+), 213 (M+H+-pyrrolidine); IR (neat): ν=2954, 2928, 2857, 2808, 1598, 1489, 1463, 1443, 1370, 1253, 1193, 1143, 1101, 1070, 1026, cm−1; HRMS (m/z):[M+H]+ calcd. for C20H30N, 284.2373; found, 284.2373.

Synthesis of allene 4j from propargylic amine

To a dried Schlenk tube was added CdI2 (293.5 mg, 0.8 mmol) inside a glove box. The Schlenk tube was then taken out and dried under vacuum with a heating gun until the white CdI2 turned to yellow green. 5j (284.1 mg, 1 mmol) and toluene (5 ml) were then added under Ar atmosphere. The Schlenk tube was then equipped with a condenser and placed in a pre-heated oil bath of 130 °C with stirring for 4 h as monitored by TLC. After cooling to room temperature, the crude reaction mixture was filtrated through a short pad of silica gel eluted with ether (20 ml). After evaporation, the residue was purified by chromatography on silica gel to afford 4j (176.9 mg, 82%) (eluent: petroleum ether) as a liquid: TLC (petroleum ether): Rf=0.94; 1H NMR (400 MHz, CDCl3) δ 7.32−7.25 (m, 4H, Ar−H), 7.20−7.12 (m, 1H, Ar−H), 6.07−6.01 (m, 1H, CH=C=C), 2.07 (td, J=7.4 Hz, 2.5 Hz, 2H, CH2), 1.80 (d, J=2.8 Hz, 3H, CH3), 1.52−1.44 (m, 2H, CH2), 1.38−1.21 (m, 6H, 3 × CH2), 0.86 (t, J=7.0 Hz, 3H, CH3); 13C NMR (100 MHz, CDCl3) δ 202.7, 136.1, 128.4, 126.5, 126.3, 103.7, 93.8, 34.1, 31.7, 29.1, 27.5, 22.7, 18.8, 14.0.

Additional information

How to cite this article: Tang, X. et al. Cadmium iodide-mediated allenylation of alkynes with ketones. Nat. Commun. 4:2450 doi: 10.1038/ncomms3450 (2013).