Nickel-Catalyzed Carbonylative Synthesis of Functionalized Alkyl Iodides.

Functionalized alkyl iodides are important compounds in organic chemistry and biology. In this communication, we developed an interesting nickel-catalyzed carbonylative synthesis of functionalized alkyl iodides from aryl iodides and ethers. With Mo(CO)6 as the solid CO source, both cyclic and acyclic ethers were activated, which is also a challenging topic in organic synthesis. Functionalized alkyl iodides were prepared in moderate to excellent yields with outstanding functional group tolerance. Besides the high value of the obtained products, all the atoms from the starting materials were incorporated in the final products and the reaction had high atom efficiency as well.

Introduction

Functionalized alkyl iodides are potent chemicals in organic chemistry and biology, and many drugs were effectively prepared from alkyl iodides (Trewick et al., 2002, DeGraw et al., 1997, Liu et al., 2014). Furthermore, ethers are widely used as solvents in organic transformations owing to their high chemical stability and relatively low boiling points. Besides, ethers are also used as strategic protecting groups for hydroxyl functions in organic synthesis (Wuts and Greene, 2014). The cleavage and functionalization of ethers is a versatile reaction in organic synthesis; several methods have been developed for the cleavage of ethers. In addition, many cyclic ethers are derived from biomass, and their ring-opening functionalization is synthetically useful for the synthesis of value-added chemicals (Christensen et al., 2014, Kumbhalkar et al., 2017, Mukherjee et al., 2017, Didiuk et al., 1995, Ohshita et al., 2006, Frei et al., 2018, Jones et al., 2017, Lübcke et al., 2017). However, owing to the high bond energy of C-O bond, relatively harsh conditions are usually needed for these transformations.

On the other hand, transition metal-catalyzed carbonylation reactions, using easily available starting materials to generate synthetically useful carbonyl-containing compounds by incorporating one or more CO into the substrate, have now emerged as one of the most powerful platform in synthetic chemistry (Gabriele et al., 2012; Wu, 2016, Peng and Wu, 2018). Typically, nucleophiles such as alcohols, amines, alkynes, and organic metallic compounds were usually used as the reactants for the carbonylative coupling reactions. Studies on carbonylative coupling reactions with weak nucleophiles (i.e., arenes and ethers) have been rarely reported. Recently, Arndtsen and co-workers developed an elegant procedure on carbonylation of arenes for the synthesis of ketones via the in situ formation of aroyl triflate as a highly reactive electrophile (Kinney et al., 2018, Tjutrins and Arndtsen, 2016, Torres et al., 2016, Quesnel and Arndtsen, 2013). However, to the best of our knowledge, the use of ether as the nucleophile in the carbonylation has not been reported because of its low reactivity (Seki et al., 1977, Tsuji et al., 1989, Watanabe et al., 1994, Getzler et al., 2004). Among the catalysts studied, palladium catalysts are more frequently applied. Although nickel catalysts have shown exceptional activities in certain bond activations, their related studies in carbonylation are still rare, which is mainly due to the fear of toxic and volatile Ni(CO)4 formation, and the situation gets even worse with the usage of CO gas. To circumvent the discussed problems, the use of CO surrogates could provide high potential (Morimoto and Kakiuchi, 2004, Wu et al., 2017, Peng et al., 2017). Herein we wish to report our new results on nickel-catalyzed carbonylative cleavage of ethers with Mo(CO)6 as the solid CO source. A broad range of functionalized alkyl iodides can be synthesized from aryl iodides and both cyclic and acyclic ethers. The control experiments showed that the carbonylative cleavage of ether was not proceeding via the intermediate acyl iodide. The ester formation step and the iodine attachment step took place simultaneously and synergistically.

Results and Discussion

Optimization Study

Initially, iodobenzene 1a and tetrahydrofuran 2b were selected as the model substrates for this carbonylative ring opening of cyclic ethers. To our delight, upon stirring a solution of iodobenzene 1a, tetrahydrofuran 2b, and Mo(CO)6 in the presence of NiCl2 and dtbbpy (4,4′-di-tert-butyl-2,2′-bipyridine) in toluene at 120°C, the carbonylative ring-opening reaction proceeded successfully and the desired product δ-iodobutyl benzoate 3ab was obtained in 52% yield (Table 1, entry 1). The reaction temperature played an important role in this reaction; increasing the reaction temperatures resulted in decreased yields (Table 1, entries 2 and 3). However, the yields decreased dramatically when the reaction was performed below 120°C (see details in Table S1). This is explainable since a high reaction temperature would facilitate the oxidative addition of transition metal into the C-I bond and the release of CO from Mo(CO)6, whereas very high reaction temperature would result in the decarbonylation of acyl nickel complex. Subsequently, different nickel catalysts such as NiBr2, Ni(acac)2, and Ni(OTf)2 were investigated (Table 1, entries 4–6, see details in Table S2). All the tested nickel catalysts showed comparable catalytic activity, with Ni(OTf)2 giving the best result and a slightly higher yield of 58% (Table 1, entry 6). No carbonylation product was obtained in the absence of nickel catalyst (see in Table S2). Then, we investigated the effect of different ligands. A decreased yield of 24% was obtained when bpy (2,2′-bipyridine) was used as the ligand (Table 1, entry 7). Other bpy-based N-ligands including phenanthroline and terpyridine failed to catalyze this transformation (see details in Table S3). In addition to N-ligands, monodentate phosphine ligands and N-heterocyclic carbene also showed catalytic activity for this reaction, albeit leading to decreased yields (Table 1, entries 8 and 9, see details in Table S3). The screening of the amount of Mo(CO)6 revealed that 0.5 equivalent of Mo(CO)6 is optimal and the product 3ab was produced in 64% yield (Table 1, entry 10). Increasing the amount of Mo(CO)6 leads to lower yields, which might be caused by the formation of catalytically nonactive Ni(CO)4 in the presence of excess CO (see details in Table S4). Screening of the additives such as NaI and TBAI did not improve the yields (see details in Table S5). Notably, the selection of solvent also played an important role in the carbonylation reactions. The yields were decreased to 57% and 42% when the reaction was performed in xylene and cyclohexane, respectively (Table 1, entries 11 and 13). However, a high yield of 89% was obtained when chlorobenzene was used as the solvent (Table 1, entry 12; see details in Table S6). This may due to the high polarity of PhCl and also the better solubility of substrates. Furthermore, increasing the amount of tetrahydrofuran improved the yield, and the desired product 3ab was isolated in 93% yield (Table 1, entry 14; see details in Table S7). In addition to Mo(CO)6, other transition metal carbonyl complexes such as Fe3(CO)12 and Cr(CO)6 were also capable of promoting this reaction, albeit giving lower yields (Table 1, entries 15 and 16). However, no desired product was obtained when gaseous CO (1 atm) was used instead of Mo(CO)6 (see Table S8). Notably, NaBr and NaCl were added into our optimized model system to attempt to produce the corresponding alkyl bromide and alkyl chloride, but no desired product could be detected.

Table 1

Optimization of the Reaction Conditions

Entry	[Ni]	Ligand	Solvent	Temperature °C)	Yield (%)a,b
1	NiCl2	dtbbpy	Toluene	120	52
2	NiCl2	dtbbpy	Toluene	130	45
3	NiCl2	dtbbpy	Toluene	140	29
4	NiBr2	dtbbpy	Toluene	120	57
5	Ni(acac)2	dtbbpy	Toluene	120	30
6	Ni(OTf)2	dtbbpy	Toluene	120	58
7	Ni(OTf)2	bpy	Toluene	120	24
8	Ni(OTf)2	PCy3	Toluene	120	48
9	Ni(OTf)2	PPh3	Toluene	120	30
10c	Ni(OTf)2	dtbbpy	Toluene	120	64
11c	Ni(OTf)2	dtbbpy	Xylene	120	57
12c	Ni(OTf)2	dtbbpy	Chlorobenzene	120	89
13c	Ni(OTf)2	dtbbpy	Cyclohexane	120	42
14c,d	Ni(OTf)2	dtbbpy	Chlorobenzene	120	95 (93)
15c,d,e	Ni(OTf)2	dtbbpy	Chlorobenzene	120	32
16c,d,f	Ni(OTf)2	dtbbpy	Chlorobenzene	120	41

See also Tables S1–S8.

Reaction conditions: iodobenzene (0.5 mmol), THF (2.0 equiv), Ni(OTf)2 (5 mol %), dtbbpy (5 mol %), Mo(CO)6 (1.0 equiv), chlorobenzene (2 mL), 120°C.

Yields were determined by gas chromatography using dodecane as an internal standard.

Mo(CO)6 (0.5 equiv).

THF (3.0 equiv).

Fe3(CO)12 (0.25 equiv).

Cr(CO)6 (0.5 equiv).

Scope of the Investigation

With the optimized conditions in hand (Table 1, entry 14), we investigated the substrate scope of this transformation. First, as summarized in Figure 1, this reaction showed good generality to aryl iodides. A series of different aryl iodides were successfully applied to the optimized reaction conditions, and the corresponding iodoester products were obtained in moderate to excellent yields. Both electron-donating group- (Figure 1, 3ab–3ib) and electron-withdrawing group- (Figure 1, 3jb–3qb) substituted iodobenzenes were well tolerated and produced the corresponding iodoester products in 42%–93% yields. The steric properties of the iodobenzene affected the yields of the reaction significantly. For example, meta- and para-substituted iodobenzenes delivered the corresponding products in moderate to good yields under the standard conditions (Figure 1, 3cb and 3db, 3nb and 3mb). On the contrary, ortho-methyl iodobenzene gave a low yield of 51% (which could be improved to 70% by increasing the reaction temperature to 130°C, Figure 1, 3bb). However, when ortho-chloro-iodobenzene was subjected to this reaction, only trace amount of the desired product was detected (Figure 1, 3ob). Notably, a range of functional groups including ketone and ester were well tolerated in this reaction (3jb and 3kb). Besides, fluoro and chloro substituents were compatible in this reaction and produced the corresponding products without breaking the C-X bonds (Figure 1, 3lb–3nb). Interestingly, when 4-bromo-iodobenzene was used as the substrate, an inseparable mixture of 3pb and 3pb′ (3.5:1) was obtained in 45% yield. However, no desired product could be detected when bromobenzene or 4-bromobenzotrifluoride was applied as the substrate, even though 1 equivalent of NaI additive was added. In addition to substituted iodobenzenes, other types of aromatic iodides were also tolerated in this transformation. For example, 1-iodonaphthalene, 2-iodonaphthalene, and 3-iodothiophene delivered the corresponding products in 77%, 72%, and 62% yields, respectively (Figure 1, 3rb, 3sb, and 3tb). In addition, bromobenzene and 1-bromo-4-(trifluoromethyl)benzene were tested under our standard conditions as well, and no conversion of the substrates could be detected.

Figure 1

Substrate Scope of Aryl Iodides

Top: reaction conditions: aryl iodide (0.5 mmol), THF (3.0 equiv), Ni(OTf)2 (5 mol %), dtbbpy (5 mol %), Mo(CO)6 (0.5 equiv), chlorobenzene (2 mL), 120°C, isolated yields.

Bottom: aryl iodide (0.5 mmol), THF (3.0 equiv), Ni(OTf)2 (5 mol %), dtbbpy (5 mol %), Mo(CO)6 (0.5 equiv), chlorobenzene (2 mL), 130°C for 3bb′ and 3sb.

See also Figures S1–S36.

Subsequently, we turned our attention to test the generality of the ethers for this carbonylative ether cleavage reaction. As illustrated in Figure 2, a series of cyclic and acyclic ethers were tolerated in this reaction system. Symmetrical cyclic ethers such as tetrahydropyran 2c and 7-oxabicyclo[2.2.1]heptane 2d were well tolerated in this reaction, and the corresponding products were produced in moderate yields (Figure 2, 3ac and 3hd). It should be mentioned that when oxetane 2a was applied to the standard reaction conditions, only trace amount of 3ha was detected. However, 3ha could be obtained in 41% yield when the reaction temperature was increased to 140°C. The low yield of 3ha might be result from the polymerization of oxetane 2a. In the case of unsymmetrical cyclic ethers, the cleavage of ether might take place via two different ways of C-O bond breaking. Generally, the C-O bond breaking prefers to take place at the sterically bulkier carbon of the ethers. For example, when 2-methyltetrahydrofuran 2f was used as the starting material, carbonylative ring-opening product 3af was obtained as the major products (Figure 2, 3af: 3af′ = 9:1). However, when 3-methyltetrahydrofuran 2g was used in this reaction, an inseparable mixture of 3ag and 3ag′ was obtained in 68% yield with low selectivity (2:1). When 2,3-dihydrobenzofuran 2e was applied to the optimized reaction condition, the carbonylative ring-opening product 3ae was obtained in 37% yields. The C-O bond cleavage occurred selectively at the C(sp3)-O bond. In addition, acyclic ethers were also explored as substrates. The carbonylative cleavage of ethyl ether 2h, butyl ether 2i, and isopropyl ether 2j took place smoothly and produced the corresponding esters in 65%, 48%, and 62% yields, respectively (Figure 2, 3ah, 3ai, and 3aj). When unsymmetrical ether 2k was subjected to this reaction, a mixture of esters were obtained with a total yield of 70% (Figure 2, 3ak: 3ak′ = 5:3). Moreover, when (methoxymethyl)benzene 2l was reacted with iodobenzene under the optimized reaction condition, only 24% yield of methyl benzoate 3al was obtained, whereas no benzyl benzoate was detected.

Figure 2

Substrate Scope of Ethers

Top: reaction conditions: aryl iodide (0.5 mmol), ether (3.0 equiv), Ni(OTf)2 (5 mol %), dtbbpy (5 mol %), Mo(CO)6 (0.5 equiv), chlorobenzene (2 mL), 120°C, isolated yields.

Middle: aryl iodide (0.5 mmol), ether (3.0 equiv), Ni(OTf)2 (5 mol %), dtbbpy (5 mol %), Mo(CO)6 (0.5 equiv), chlorobenzene (2 mL) at 130°C for 3ha.

Bottom: aryl iodide (0.5 mmol), ether (3.0 equiv), Ni(OTf)2 (5 mol %), dtbbpy (5 mol %), Mo(CO)6 (0.5 equiv), chlorobenzene (2 mL) at 140°C for 3ha′.

See also Figures S37–S58.

Mechanistic Study

To gain a better understanding of the reaction pathway of this carbonylative ether cleavage reaction, a series of supporting experiments were conducted (Scheme 1). First, when iodobenzene 1a and THF (tetrahydrofuran) 2b were subjected to the standard conditions in the presence of excess halide additives such as LiCl and LiBr, the iodoester 3ab was obtained in good yields. No chloro- or bromoesters were observed (Scheme 1, A). Besides, 1,4-dioxane was reported to give the chloroester in acylative cleavage reactions. However, when 1,4-dioxane 2m was used in this reaction, instead of iodoester 3am, ethylene glycol dibenzoate 3am′ (Figures S59 and S60) was obtained in 40% yield (Scheme 1, B). This might result from the coordination of 3am to the nickel catalyst, thus facilitating a second carbonylative cleavage. Furthermore, when unsymmetrical cyclic ether 2f was reacted with iodobenzene, 3af and 3af′ were obtained in a ratio of 9:1. However, the reaction of in situ prepared benzoyl iodide with 2-methyltetrahydrofuran 2f produced 3af and 3af′ with an opposite ratio of 1:2 (Scheme 1, C). These phenomena revealed that the carbonylative cleavage of ether did not proceed via the intermediate acyl iodide. The ester formation step and the iodine attachment step should take place simultaneously and synergistically.

Scheme 1

Mechanistic Studies with Fuans and 1,4-Dioxane

Plausible Reaction Mechanism

Based on these results and previous literature, a plausible mechanism was proposed in Scheme 2. Initially, the oxidative addition of aryl iodide to the in situ-generated Ni(0) formed the aryl nickel complex 4. Then, coordination and insertion of CO, which was released from Mo(CO)6, generated the acyl nickel complex 5. Then, the coordination of the oxygen of THF to nickel activated the oxygen and the adjacent carbon. Subsequently, the iodide attacked the carbon and the C-O bond broke to generate intermediate 7. Finally, reductive elimination of 7 delivered the desired product 3 and regenerated Ni(0) for the next catalytic cycle.

Scheme 2

A Plausible Mechanism

Conclusion

In summary, we have developed an interesting nickel-catalyzed carbonylative cleavage of ethers with Mo(CO)6 as the solid CO source. Functionalized alkyl iodides were produced in moderate to good yields from aryl iodides and both cyclic and acyclic ethers. Beside the high value of the obtained products, all atoms from the starting materials were incorporated in the final products.

Limitations of Study

Substrates such as aryl bromides and aryl chlorides failed in this system. Alkyl bromides and alkyl chlorides cannot be prepared. More detailed mechanistic studies still need to done.

Methods

All methods can be found in the accompanying Transparent Methods supplemental file.