Pd-Catalyzed coupling reaction of cyclobutanols with propargylic carbonates

Abstract

An efficient approach of Pd-catalyzed ring opening coupling reaction of cyclobutanols with propargylic carbonates was realized to provide a series of multisubstituted δ-allenyl ketones. The reaction had a wide substrate scope with tolerance to different functional groups.

---

Introduction

An allene unit is an important type of molecular structure and found widespread in numerous natural products and bioactive molecules.¹ Thus, the synthesis of differently substituted allenes has attracted a lot of attention.² The coupling reaction of propargylic carbonates with different metal regents was an efficient method for the construction of multi-substituted allenes.³ Recently, we found that cyclopropanols may be used as an efficient coupling reagent with propargylic carbonates for the synthesis of different substituted allenes. The reaction has a wide substrate scope under mild reaction conditions.⁴ To expand the scope of such a reaction, we envisioned that cyclobutanols which are easily available from cyclobutanones⁵ could also be used in this type of coupling reaction via Pd catalysts as pioneered by Uemura.^6,7 In this paper, we report a general method for the synthesis of different substituted δ-allenyl ketones from propargylic carbonates.

Firstly, standard reaction conditions with cyclopropanols were used for the coupling reaction of cyclobutanol 1a with propargylic carbonate 2a. Unfortunately, the reaction mixture was stirred at 50 °C for 21 h and then 80 °C for 7 h to afford the desired allene product 3aa in only 28% NMR yield with 46% of carbonate 2a being recovered (Scheme 1). These results revealed that the reactivity of cyclobutanol was much lower than the cyclopropanol, which was possibly due to their different ring strains (the ring strains of cyclopropane and cyclobutane are 29.0 and 26.3 kcal mol⁻¹, respectively⁸).

Scheme 1 Reaction of cyclobutanol 1a with propargylic carbonate 2a under the standard conditions of cyclopropanol.

Results and discussion

Based on these results, we decided to further screen the reaction conditions with cyclobutanol 1a and propargylic carbonate 2a as the model substrates. Firstly, different carbonate salts were added in toluene at 80 °C with Pd(dba)₂ and XPhos as the catalysts. K₂CO₃ and Cs₂CO₃ provided the highest yields (Table 1, entries 1–4). K₃PO₄·3H₂O had almost the same effect (Table 1, entry 5). Other bases such as t-BuOK, NaHCO₃, CsOPiv, DIPEA, and Et₃N were also tested demonstrating inferior results (Table 1, entries 6–10). Finally, we also found that the very freshly prepared carbonate 2a can be used directly to give the corresponding allene 3aa in 80% NMR yield without the addition of any base (Table 1, entry 11). However, 1.0 equiv. of base was required for the (slightly) aged 2a.

Table 1 The effect of base^a

Entry	Base	NMR yield of 3aa ^b (%)	Recovery of 2a ^b (%)
a Reaction conditions: 1a (0.24 mmol), 2a (0.20 mmol), Pd(dba)2 (2.5 mol%), XPhos (5.0 mol%), and base (1.0 equiv.) in toluene (1.0 mL) at 80 °C. b NMR yield with CH2Br2 as the internal standard. c The carbonate 2a was used immediately after chromatography on silica gel; reaction time: 3 h.
1	Li2CO3	59	3
2	Na2CO3	54	10
3	K2CO3	80	0
4	Cs2CO3	80	0
5	K3PO4·3H2O	79	0
6	t-BuOK	Complicated	—
7	NaHCO3	55	12
8	CsOPiv	0	96
9	DIPEA	51	12
10	Et3N	61	3
11^c	—	80	0

---

Then, a range of different phosphine ligands were screened. The use of PPh₃ and dppp led to complicated results (Table 2, entries 1 and 2). When other phosphine ligands such as PCy₃, dppf, dppm, LB-Phos and Gorlos-Phos were used, the starting material 2a was recovered in 36–75% with no product 3aa being observed (Table 2, entries 3–7). Fortunately, SPhos and rac-BINAP resulted in a moderate yield of product 3aa with 5–10% yield of 2a remaining (Table 2, entries 8 and 9). Finally, it was observed that XPhos was still the best ligand, which provided the allene product in 80% yield and no 2a was recovered (Table 2, entry 10).

Table 2 The effect of ligand^a

Entry	Ligand	NMR yield of 3aa ^b (%)	Recovery of 2a ^b (%)
a Reaction conditions: 1a (0.24 mmol), 2a (0.20 mmol), Pd(dba)2 (2.5 mol%), ligand (5.0 mol%), and K2CO3 (1.0 equiv.) in toluene (1.0 mL) at 80 °C. b NMR yield with CH2Br2 as the internal standard.
1	PPh3	Complicated	—
2	dppp	Complicated	—
3	PCy3	0	75
4	dppf	0	74
5	dppm	0	36
6	LB-Phos	0	63
7	Gorlos-Phos	0	39
8	SPhos	56	10
9	rac-BINAP	63	5
10	XPhos	80	0

---

Subsequently, with XPhos as the optimal ligand, Pd(II) salts were tested. Different Pd(II) sources such as Pd(OAc)₂, PdCl₂ and Pd(TFA)₂ gave 9–28% yields of product 3aa with 31–81% recovery of the starting material 2a (Table 3, entries 2–4).

Table 3 The effect of Pd-catalyst^a

Entry	[Pd]	NMR yield of 3aa ^b (%)	Recovery of 2a ^b (%)
a Reaction conditions: 1a (0.24 mmol), 2a (0.20 mmol), [Pd] (2.5 mol%), XPhos (5.0 mol%), and K2CO3 (1.0 equiv.) in toluene (1.0 mL) at 80 °C. b NMR yield with CH2Br2 as the internal standard.
1	Pd(dba) 2	80	0
2	Pd(OAc)2	28	31
3	PdCl2	13	73
4	Pd(TFA)2	9	81

---

Considering the convenience, K₃PO₄·3H₂O was used as the base for further study. Solvents were further screened. THF, dioxane, CH₃CN and DCM all gave the desired product 3aa albeit with lower yields (49–69%, Table 4, entries 2–5). When DMF was used, the reaction was complicated and 14% of 3aa was observed (Table 4, entry 6).

Table 4 The effect of solvent^a

Entry	Solvent	NMR yield of 3aa ^b (%)	Recovery of 2a ^b (%)
a Reaction conditions: 1a (0.24 mmol), 2a (0.20 mmol), Pd(dba)2 (2.5 mol%), XPhos (5.0 mol%), and K2CO3 (1.0 equiv.) in solvent (1.0 mL) at 80 °C. b NMR yield with CH2Br2 as the internal standard.
1	Toluene	79	0
2	THF	69	0
3	Dioxane	61	0
4	CH3CN	56	0
5	DCM	49	0
6	DMF	14	—

---

Finally, lowering the temperature to 70 °C or 60 °C can both provide 81% yield of 3aa (Table 5, entries 2 and 3). However, the reaction at 60 °C on 1 mmol scale led to the minor recovery of the propargylic carbonate 2a. Decreasing the temperature further to 50 °C resulted in a lower yield of 3aa with 53% recovery of the starting material 2a (Table 5, entry 4).

Table 5 The effect of temperature^a

Entry	Temperature	NMR yield of 3aa ^b (%)	Recovery of 2a ^b (%)
a Reaction conditions: 1a (0.24 mmol), 2a (0.20 mmol), Pd(dba)2 (2.5 mol%), XPhos (5.0 mol%), and K2CO3 (1.0 equiv.) in toluene (1.0 mL). b NMR yield with CH2Br2 as the internal standard.
1	80	79	0
2	70	81	0
3	60	81	0
4	50	31	53

---

With the optimized reaction conditions above (Table 5, entry 2), the substrate scope of various cyclobutanols 1 with the propargylic carbonates 2a was investigated. All the reactions were carried out on 1 mmol scale and the results are summarized in Scheme 2. When R¹ is an aryl group, besides phenyl, the substituent group on the phenyl group can be either an electron-withdrawing group (F, Cl) or an electron-donating group (OMe), providing the allene products in 50–79% yields (Scheme 2, 3aa–3ga). In addition, the substituents on the phenyl rings at the ortho, meta, and para positions may all be tolerated (Scheme 2, 3da–3fa). And the piperonyl group can be tolerated well providing 55% yield of 3ga. Except for the aryl group, R¹ can be different alkyl groups such as n-butyl, cyclohexyl and benzyl to provide the desired allene products in 65–72% yields (Scheme 2, 3ha–3ja). The reaction also worked with a cyclobutanol with an additional substituent group at the 3-position providing the expected product 3ka with 40% yield (Scheme 2).

Scheme 2 Substrate scope with respect to cyclobutanols. Reaction conditions: 1 (1.2 mmol), 2a (1.0 mmol), Pd(dba)₂ (2.5 mol%), XPhos (5.0 mol%), and K₃PO₄·3H₂O (1.0 equiv.) in toluene (5.0 mL) at 70 °C. ^a Pd(dba)₂ (5.0 mol%) and XPhos (10.0 mol%) were used. ^b Byproduct, 1,3-dipenylpropadiene, was formed in 4% NMR yield.

Notably, when benzocyclobutanol 1l was used in the reaction, product 3la with the cleavage of the proximal bond was selectively generated in 51% yield, and the distal bond cleavage product 3la′ was not observed (Scheme 3).

Scheme 3 The reaction of benzocyclobutanol. Reaction conditions: 1l (1.2 mmol), 2a (1.0 mmol), Pd(dba)₂ (2.5 mol%), XPhos (5.0 mol%), and K₃PO₄·3H₂O (1.0 equiv.) in toluene (5.0 mL) at 70 °C.

However, when R¹ = H, the reaction did not provide the expected product 3ma, instead, by-products 1,3-diphenylpropadiene 3ma′ and 1,3-dipenylpropyne 3ma′′ were formed in 25% and 27% NMR yields, respectively.⁴ The reaction of the cyclobutanol 1n, which has an additional substituent at the 2-position, gave no allene product with 60% recovery of the starting material 2a (Scheme 4).

Scheme 4 Failed substrates of cyclobutanols.

Next, the scope of propargylic carbonates 2 was also tested. When R⁵ is H, R³ and R⁴ substituents can be different aryl or alkyl groups. For example, 4-acetyl, 2-chloro, and 2-methoxycarbonyl substituted phenyls at the R³ position were tolerated providing the allene products 3ab–3ad in 45–72% yields. When R³ was n-hexyl, it could result in 40% yield of 3ae by decreasing the amount of 1a to 1.0 mmol and increasing the amount of propargylic carbonate 2e to 1.5 mmol. And 4-chloro, 3,4-dichloro, 2-cyano and 2-methoxy substituted phenyls could be introduced successfully to the R⁴ position to give the final products 3af–3ai in 61–80% yields. It should be noted that the catalyst loading should be improved to 5 mol% for 3ac, 3ad, 3ag and 3ah to make sure the complete consumption of the starting carbonates. Notably, besides the secondary propargylic carbonates, tertiary propargylic carbonates (R² = Ph, R³ = ⁿPr, R⁴ = Me) could also be used to afford the fully substituted allene product 3aj in a moderate yield. Besides, when R⁴ = R⁵ = H, the reaction afforded 3ak with 50% yield. However, when R³ = H, the corresponding allene product 3al was formed in a very low yield (<5% by ¹H NMR) (Scheme 5).

Scheme 5 Substrate scope with respect to propargylic carbonates. Reaction conditions: 1a (1.2 mmol), 2 (1.0 mmol), Pd(dba)₂ (2.5 mol%), XPhos (5.0 mol%), and K₃PO₄·3H₂O (1.0 equiv.) in toluene (5.0 mL) at 70 °C. ^a 1a (1.0 mmol), 2 (1.2 mmol), Pd(dba)₂ (5.0 mol%) and XPhos (10.0 mol%) were used. ^b Pd(dba)₂ (5.0 mol%) and XPhos (10.0 mol%) were used. ^c 1a (1.0 mmol) and 2 (1.5 mmol) were used.

In order to demonstrate the synthetic utility and practicality of the method, the allene product 3ab was synthesized on a gram scale. A 5.0 mmol scale reaction was carried out providing 1.2305 g of the desired product 3ab in 66% isolated yield. And the allene product 3ab can be subjected to some transformations. For example, it can be reduced to the secondary alcohol 4 with NaBH₄, and reacted with the Wittig reagent producing product 5 (Scheme 6).

Scheme 6 Gram-scale reaction and synthetic utility.

Based on the previous experimental results and knowledge,^4,9 a possible mechanism was proposed as shown in Scheme 7. Firstly, the propargylic carbonate 2 reacted with Pd(0) via S_N2′-type oxidative addition together with the elimination of a molecule of CO₂, which leads to the allenyl palladium methoxide intermediate I. Then intermediate I underwent ligand exchange with cyclobutanol 1 generating intermediate II, which would provide intermediate III by β-C elimination. Intermediate III delivered the final product 3via reductive elimination and regenerated the Pd(0) species completing the catalytic cycle.

Scheme 7 The proposed mechanism.

Conclusions

In summary, an efficient method to synthesize different multi-substituted δ-allenyl ketones via palladium-catalyzed ring opening of cyclobutanols and oxidative addition of propargylic carbonates was developed. This reaction had a wide substrate scope with tolerance to different functional groups, and this method may be easily expanded to gram scale synthesis with potential synthetic utility.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

Financial support from the National Basic Research Program of China (2015CB856600) and the National Natural Science Foundation of China (21502020) is greatly appreciated. We thank Yifan Cui of our group for reproducing the results of 3ca, 3ga, and 3ab represented in this study.

Notes and references

1. For a review on allenes in natural products and bioactive molecules, see: A. Hoffmann-Röder and N. Krause, Angew. Chem., Int. Ed., 2004, 43, 1196.
2. For reviews on the synthesis of different substituted allenes, see: (a) A. Hoffmann-Röder and N. Krause, Angew. Chem., Int. Ed., 2002, 41, 2933; (b) N. Krause and A. Hoffmann-Röder, Tetrahedron, 2004, 60, 11671; (c) G. B. Hammond, ACS Symp. Ser., 2005, 911, 204; (d) K. M. Brummond and J. E. DeForrest, Synthesis, 2007, 795; (e) M. Ogasawara, Tetrahedron: Asymmetry, 2009, 20, 259; (f) S. Yu and S. Ma, Chem. Commun., 2011, 47, 5384; (g) R. K. Neff and D. E. Frantz, ACS Catal., 2014, 4, 519; (h) J. Ye and S. Ma, Org. Chem. Front., 2014, 1, 1210; (i) R. K. Neff and D. E. Frantz, Tetrahedron, 2015, 71, 7.
3. Selected examples, see: (a) E. Keinan and E. Bosch, J. Org. Chem., 1986, 51, 4006; (b) P. Dixneuf, T. Guyot, M. Ness and S. Roberts, Chem. Commun., 1997, 2083; (c) R. Riveiros, D. Rodriguez, J. Sestelo and L. Sarandeses, Org. Lett., 2006, 8, 1403; (d) H. Ito, Y. Sasaki and M. Sawamura, J. Am. Chem. Soc., 2008, 130, 15774; (e) H. Ohmiya, H. Ito and M. Sawamura, Org. Lett., 2009, 11, 5618; (f) Q. Li, J. Jeng and H. Gau, Eur. J. Org. Chem., 2014, 7916; (g) T. Zhao, Y. Yang, T. Lessing and K. Szabó, J. Am. Chem. Soc., 2014, 136, 7563; (h) H. Luo, Y. Yu and S. Ma, Org. Chem. Front., 2016, 3, 1705; (i) S. Kessler and J. Bäckvall, Angew. Chem., Int. Ed., 2016, 55, 3734; (j) Q. Lu, S. Greßies, F. J. R. Klauck and F. Glorius, Angew. Chem., Int. Ed., 2017, 56, 6660.
4. P. Wu, M. Jia, W. Lin and S. Ma, Org. Lett., 2018, 20, 554.
5. (a) K. Jia, F. Zhang, H. Huang and Y. Chen, J. Am. Chem. Soc., 2016, 138, 1514; (b) B. Casey, C. Eakin and R. Flowers II, Tetrahedron Lett., 2009, 50, 1264; (c) H. Zeng, P. Pan, J. Chen, H. Gong and C. Li, Eur. J. Org. Chem., 2017, 1070.
6. For examples on the coupling reaction of the ring-opening of cyclobutanols via Pd catalysis, see: (a) T. Nishimura, K. Ohe and S. Uemura, J. Am. Chem. Soc., 1999, 121, 2645; (b) T. Nishimura and S. Uemura, J. Am. Chem. Soc., 1999, 121, 11010; (c) T. Nishimura, K. Ohe and S. Uemura, J. Org. Chem., 2001, 66, 1455; (d) T. Nishimura, S. Matsumura, Y. Maeda and S. Uemura, Chem. Commun., 2002, 50; (e) T. Nishimura, S. Matsumura, Y. Maeda and S. Uemura, Tetrahedron Lett., 2002, 43, 3037; (f) M. Ethirajan, H. Oh and J. Cha, Org. Lett., 2007, 9, 2693; (g) A. Chtchemelinine, D. Rosa and A. Orellana, J. Org. Chem., 2011, 76, 9157; (h) A. Ziadi and R. Martin, Org. Lett., 2012, 14, 1266; (i) A. Ziadi, A. Correa and R. Martin, Chem. Commun., 2013, 49, 4286; (j) L. Chen, F. Sun, Y. Sun, Z. Xu, Z. Zheng, Y. Cui, J. Cao and L. Xu, Adv. Synth. Catal., 2018, 360, 411.
7. S. Matsumura, Y. Maeda, T. Nishimura and S. Uemura, J. Am. Chem. Soc., 2003, 125, 8862.
8. P. Khoury, W. Tam and J. Goddard, Tetrahedron, 2004, 60, 8103.
9. For a review, see: (a) J. Tsuji and T. Mandai, Angew. Chem., Int. Ed. Engl., 1995, 34, 2589. See also: (b) S. Ogoshi, K. Tsutsumi, M. Ooi and H. Kurosawa, J. Am. Chem. Soc., 1995, 117, 10415; (c) T. Konno, M. Tanikawa, T. Ishihara and H. Yamanaka, Chem. Lett., 2000, 1360; (d) B. M. Trost and L. Debien, Chem. Sci., 2016, 7, 4985.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9qo00192a

---