Decarbonylative Sonogashira Cross-Coupling of Carboxylic Acids

Abstract

Decarbonylative Sonogashira cross-coupling of carboxylic acids by palladium catalysis is presented. The carboxylic acid is activated in situ by the formation of a mixed anhydride, further decarbonylates using Pd(OAc)₂/Xantphos system to provide aryl–Pd intermediate, which is intercepted by alkynes to access the traditional Pd(0)/(II) cycle using carboxylic acids as ubiquitous and orthogonal electrophilic cross-coupling partners. The methodology efficiently constructs new C(sp²)–C(sp) bonds and can be applied to the derivatization of pharmaceuticals. Mechanistic studies give support to decarbonylation preceding transmetallation in this process.

Graphical Abstract

Sonogashira cross-coupling represents a powerful and one of the most widely utilized strategies for incorporating alkynes into organic molecules.¹ Recent surveys place Sonogashira cross-coupling as the third most commonly used cross-coupling method in industrial settings.² In particular, the capacity of alkynes to function as an important functional group on their own as well as participate in an array of reactions as synthetic intermediates³ has resulted in a broad interest to establish new protocols to incorporate alkyne motifs by the Sonogashira cross-coupling.^1,2

Typical electrophilic substrates for Sonogashira cross-coupling include aryl halides and pseudohalides, such as sulfonates,^1,2 while more recent attention has focused on iodonium salts, diazonium salts, phosphonium salts, and sulfonium salts among other electrophiles by C–X, C–O, C–N, C–P and C–S cleavage.⁴ In this context, it is important to note that while many exciting achievements have been made using Pd/Cu-co-catalyzed variant, including cross-coupling of alkyl electrophiles,⁵ the Pd-catalyzed/Cu-free variant has received increased attention.^6a–c Recently, C–H alkynylation methods, in particular transforms enabled by directing groups have also been developed.⁷

In this context, our laboratory has developed new methods for decarbonylative cross-coupling of carboxylic acids.^8,9 In this reaction manifold, carboxylic acid is activated in situ to form a mixed anhydride,^8a–e followed by selective oxidative addition of the C(O)–OR bond to a low valent metal and decarbonylation to furnish the traditional Ar–metal intermediate using ubiquitous carboxylic acids as cross-coupling partners.¹⁰ The particular value of this approach is that carboxylic acids represent pervasive substrates in organic synthesis¹¹ and are derived from an orthogonal pool of precursors to aryl halides and pseudohalides.¹¹ The inherent presence of carboxylic acids in pharmaceuticals, natural products and functional materials renders the growing cross-coupling repertoire of carboxylic acids attractive for organic synthesis.¹²

In the continuation of our studies on decarbonylative borylation,^8a arylation,^8b reduction,^8c phosphorylation^8d and heteroarylation^8e of carboxylic acids,⁹ we became interested in the development of decarbonylative Sonogashira cross-coupling of carboxylic acids (Figure 1). Herein, we present the development of this method, including derivatization of pharmaceuticals and mechanistic studies that give support to decarbonylation preceding transmetallation in this process. The identified catalyst system involves Pd(OAc)₂ (5 mol%)/Xantphos (10 mol%) with piv₂O (1.5 equiv) and DMAP (1.5 equiv), dioxane (0.25 M), 160 °C, 15 h, as activators. Very recently, Chen and co-workers reported the first study of the decarbonylative cross-coupling of carboxylic acids with alkynes.¹³ The conditions reported use Pd₂(dba)₃ (2.5 mol%), Xantphos (10 mol%), Ac₂O (1.5 equiv), DME (0.10 M), 130 °C, 12 h.¹³ In our experience^8a–e piv₂O with or without Lewis base is superior to Ac₂O in promoting selective decarbonylative coupling of carboxylic acids.^8,10,11 In light of this development and our own studies, we considered it appropriate to report our findings. The two catalytic systems should be considered complementary,¹⁴ while our study excludes the direct decarbonylation of ynones as intermediates in this process.

Figure 1.

(a) Sonogashira cross-coupling. (b) This work: Sonogashira cross-coupling of carboxylic acids by decarbonylation.

Selected optimization results are presented in Table 1. After extensive optimization, we identified the combined use of Pd(OAc)₂ and Xantphos in dioxane at 160 °C as effective catalytic promoters for the process (entries 1–19). Interestingly, several phosphine ligands can be used, including dppb, dppp, dppent, dppf; however, Xantphos proved most effective, while monodentate phosphines were ineffective (entries 10–19). Out of several activators tested, such as Ac₂O, Boc₂O and piv₂O, the latter was identified as giving the best reactivity (entries 20–22), in agreement with our previous studies.^8,14 It is important to note that Cu is not required for this process (entries 24–31), resulting in a Cu-free Sonogashira variant. Finally, we demonstrated that the reaction proceeds at temperatures as low as 120 °C, demonstrating efficient decarbonylation under these conditions (entry 33). Furthermore, the use of Pd(OAc)₂ at 1 mol% resulted in promising 45% yield for future reaction development (entry 35). Overall, the catalytic system is complementary to the one developed by Chen and co-workers.¹³ The use of large bite angle phosphines, such as Xantphos (108°), promotes decarbonylation in the process.^8a–e

Table 1.

Optimization of the Cross-Coupling^a

entry	[Pd]	ligand	base	additive	yield (%)
1	-	-	DMAP	piv2O	<2
2b	-	-	DMAP	piv2O	<2
3	Pd(OAc)2	dppb	DMAP	piv2O	68
4	Pd(OAc)2	dppb	-	piv2O	7
5	Pd(OAc)2	dppb	DMAP	-	<2
6	Pd(OAc)2	dppb	Et3N	piv2O	18
7	Pd(OAc)2	dppb	py	piv2O	<2
8	Pd2(dba)3	dppb	Na2CO3	piv2O	<2
9	Pd(OAc)2	dppb	K2CO3	piv2O	33
10	Pd(OAc)2	dppp	DMAP	piv2O	45
11	Pd(OAc)2	dppent	DMAP	piv2O	35
12	Pd(OAc)2	dppf	DMAP	piv2O	31
13	Pd(OAc)2	XantPhos	DMAP	piv2O	92
14	Pd(OAc)2	DPEPhos	DMAP	piv2O	7
15	Pd(OAc)2	BINAP	DMAP	piv2O	<2
16	Pd(OAc)2	PCy3	DMAP	piv2O	<2
17	Pd(OAc)2	PCy2Ph	DMAP	piv2O	<2
18	Pd(OAc)2	PPh3	DMAP	piv2O	<2
19	Pd(OAc)2	DavePhos	DMAP	piv2O	<2
20	Pd(OAc)2	XantPhos	DMAP	Ac2O	47
21	Pd(OAc)2	XantPhos	DMAP	Boc2O	58
22c	Pd(OAc)2	XantPhos	DMAP	piv2O	74
23d	Pd(OAc)2	XantPhos	DMAP	piv2O	80
24e	Pd(OAc)2	XantPhos	DMAP	piv2O	13
25f	Pd(OAc)2	XantPhos	DMAP	piv2O	77
26g	Pd(OAc)2	XantPhos	DMAP	piv2O	87
27h	Pd(OAc)2	XantPhos	DMAP	piv2O	30
28i	Pd(OAc)2	XantPhos	DMAP	piv2O	42
29j	Pd(OAc)2	XantPhos	DMAP	piv2O	37
30k	Pd(OAc)2	XantPhos	DMAP	piv2O	24
31l	Pd(OAc)2	XantPhos	DMAP	piv2O	3
32m	Pd(OAc)2	XantPhos	DMAP	piv2O	84
33n	Pd(OAc)2	XantPhos	DMAP	piv2O	69
34o	Pd(OAc)2	XantPhos	DMAP	piv2O	73
35p	Pd(OAc)2	XantPhos	DMAP	piv2O	45

^aConditions: 1-Np-CO₂H (1.0 equiv), alkyne (4.0 equiv), Pd(OAc)₂ (5 mol%), ligand (10 mol%), DMAP (1.5 equiv), piv₂O (1.5 equiv), dioxane (0.25 M), 160 °C, 15 h.

^bCuI (10 mol%).

^calkyne (3.0 equiv).

^dalkyne (5.0 equiv).

^eCuCl (10 mol%).

^fCuBr (10 mol%).

^gCuI (10 mol%).

^hCuCN (10 mol%).

ⁱCuF₂ (10 mol%).

^jCuSO₄ (10 mol%).

^kCu(OAc)₂ (10 mol%).

^lCu(OTf)₂ (10 mol%).

^m140 °C.

ⁿ120 °C.

^otoluene.

^pPd(OAc)₂ (1 mol%), Xantphos (2 mol%).

See SI for details.

Having identified optimal reaction conditions for the coupling, we investigated the scope of this decarbonylative process for the synthesis of alkynes (Scheme 1). As shown in Scheme 1A, this method is successful with an array of aryl carboxylic acids. Naphthyl-carboxylic acids (3a-3c) as well as electronically-differentiated benzoic acids (3d-3f) are well-tolerated. Importantly, halides, such as chlorides, are compatible with this process (3g), enabling derivatization by standard cross-coupling technologies and showing complementarity of our catalytic system. As shown previously by us,^8a–e the reactivity of carboxylic acids is comparable to Ar–Br in this manifold. Furthermore, steric ortho-substitution, such as Me (3h), CF₃ (3i), and even sterically-hindered Ph (3j) is also possible. As expected, meta-substitution is well-tolerated (3k-3l). This method can also be used to cross-couple heterocyclic carboxylic acids, such as thienyl-carboxylic acids (3m-3n).

Scheme 1. Decarbonylative Sonogashira Cross-Coupling of Carboxylic Acids^a,b.

^aConditions: carboxylic acid (1.0 equiv), alkyne (4.0 equiv), Pd(OAc)₂ (5 mol%), XantPhos (10 mol%), DMAP (1.5 equiv), piv₂O (1.5 equiv), dioxane (0.20 M), 160 °C, 15 h. ^bIsolated yields. See SI for details.

Finally, the potential of the method in derivatization of pharmaceuticals has been demonstrated in the direct Sonogashira cross-coupling of Bexarotene, an antineoplastic agent (3o). It is important to note that in contrast to oxidative methods for cross-coupling of carboxylic acids,^10,11 this redox-neutral manifold by decarbonylation does not require steric- or electronic bias to facilitate decarboxylation, resulting in a general method.

Next, the scope of the alkyne component was briefly investigated (Scheme 1B). As shown, the reaction is compatible with electronically-differentiated phenylacetylenes (3d’, 3p-3q). Furthermore, alkylacetylenes, such as cyclohexylacetylene, couple in high yields (3r). Sensitive alkyl electrophiles that can be utilized in further functionalization, such as halides, are also tolerated (3s). Finally, silyl-acetylenes, such as (triisopropylsilyl)acetylene can be used (3t), providing access to terminal alkynes after deprotection. At the present stage, heterocyclic acids, such as thienyl are tolerated. 3-Pyridyl carboxylic acid gave lower but promising yield (20%). Carbonyl groups, such as ketones, are tolerated in decarbonylative coupling.^8–12

Mechanistic studies were conducted to gain insight into this process (Scheme 2). Thus, intermolecular competition experiments between differently substituted acid electrophiles revealed that electron-deficient carboxylic acids are inherently more reactive (Scheme 2A). Furthermore, sterically-hindered acid electrophiles are more reactive than unsubstituted benzoic acids (Scheme 2B), which is consistent with decarbonylation favored by steric-demand of acyl–Pd complexes.¹⁰ In contrast, intermolecular competition experiments using differently substituted acetylenes revealed that electron-rich acetylenes are more reactive (Scheme 2C). Next, we prepared the mixed acyl anhydride 4 and subjected this compound to the reaction conditions, resulting in the formation of the decarbonylative cross-coupling product in 82% yield (Scheme 3A). Furthermore, to test the potential of the direct decarbonylation of ynones,¹⁰ compounds 5 and 6 were tested as potential substrates, resulting in no conversion to the desired products (Scheme 3B). We also note that subjecting phenylacetylene to the reaction conditions results in 29% of diphenylacetylene by alkyne dimerization as a minor reaction pathway. Overall, these studies are consistent with decarbonylation preceding transmetallation in the process.

Scheme 2.

Mechanistic Studies

Scheme 3.

Control Experiments

In summary, we have developed palladium-catalyzed decarbonylative Sonogashira cross-coupling of carboxylic acids. The reaction is promoted by Pd(OAc)₂/Xantphos catalytic system and proceeds by a selective oxidative addition of a mixed anhydride and decarbonylation. The reaction is compatible with various carboxylic acids and alkynes, leading to the facile formation of C(sp²)–C(sp) bonds from ubiquitous carboxylic acids. The potential of this method in pharmaceutical derivatization has been demonstrated. Mechanistic studies support the pathway by direct decarbonylation/transmetallation vs. acyl cross-coupling/ynone decarbonylation. The method expands the portfolio of decarbonylative transformations of carboxylic acids as effective aryl electrophiles in organic synthesis. Further studies on decarbonylative cross-coupling of carboxylic acids are ongoing and will be reported in due course.