Jesús Campos‡
a,
Sarah K. Goforth‡b,
Robert H. Crabtree*a and
T. Brent Gunnoe*b
aDepartment of Chemistry, Yale University, 225 Prospect Street, New Haven, Connecticut 06520, USA. E-mail: robert.crabtree@yale.edu
bDepartment of Chemistry, University of Virginia, Charlottesville, Virginia 22904-4319, USA. E-mail: tbg7h@virginia.edu
First published on 17th September 2014
A selective protocol for the metal-free α-C–H amidation of ethers using sulfonamides and hypervalent iodine oxidants has been developed. The absence of precious metals and the conditions employed make the method environmentally attractive. A number of cyclic and acyclic, linear and branched ethers have been successfully amidated, and a broad sulfonamide scope has been demonstrated. Two unusual reactions, namely the amidation of an unactivated tert-butyl group and a tandem C–C coupling reaction, are also described.
A variety of amine and oxidant combinations have been used for metal-free amination, amidation and imidation of C–H bonds. The oxidants in these systems include peroxides (tert-butyl hydroperoxide,15–18 di-tert-butyl peroxide,19 H2O220, benzoquinones,21 chloramine-T,22 and hypervalent halogen species.23–35 Additional organic catalysts (e.g., I2
15,24,29 and n-Bu4NI)17,18 and stoichiometric additives (e.g., I2,16,22,35–37 N-iodosuccinimide,16,20 nBu4NI,16 KI,16,28 TsOH19 and AcOH15,20) are often involved and perform various functions. Hypervalent iodine species, particularly iodosobenzene derivatives such as PhI(OAc)2, are prevalent in both metal-catalyzed and metal-free processes for C–H amidation.1–10,23–32,35 Owing to their high oxidation potential, these species are documented to activate sulfonamides for in situ generation of iminoiodinanes (PhI = NSO2R), thus expanding the scope of the prior strategies based on preformed iminoiodinanes (Scheme 1).7–10
While a majority of the metal-free processes for C–H amidation and amination have been applied to oxidative coupling of N–H bonds to sp and sp2 C–H bonds of unsaturated alkynes,38 arenes,23,35,37,39,40 alkenes,25–27,41 and azoles,20,31,32,42,43 several examples of sp3 C–H amidation of benzylic16–19,22–24,41 and allylic18,21,30 bonds have also been reported. In addition, Muñiz and coworkers described the amidation of an α-methyl group of a ketone.25 Metal-free amidation is rare for sp3 C–H bonds unactivated by adjacent double bonds. The I2-catalyzed oxidative coupling of purines with tetrahydrofuran (THF) oxidized by PhI(OAc)2 is known,29 and a similar system utilizing PhI(OAc)2 as oxidant but involving stoichiometric I2 has been reported for intramolecular amidation of alkyl sulfonamides.36 Most prominently, Ochiai and coworkers discovered a highly active hypervalent bromine reagent, p-(CF3)(Ph)BrNTf (Tf = CF3SO2), for α-C–H amidation of alkyl ethers (eqn (1))34 and also, in a striking example, for regioselective amidation of unfunctionalized alkanes.33
![]() | (1) |
The present work began with an initial finding that reaction of 2,2,2-trichloroethoxysulfonamide in neat THF in the presence of PhI(OAc)2 selectively produces the α-C–H amidation product (eqn (2)). The simple sulfonamide/PhI(OAc)2 system offers some advantages over the specialized p-(CF3)(Ph)BrNTf reagent reported for amidation of alkyl ethers with sulfonamides.34 The air- and moisture-stable and commercially available PhI(OAc)2 is more convenient than p-(CF3)(Ph)BrNTf, which must be handled under inert atmosphere and whose two-step synthesis involves BrF3 and liberates HF.41,44 Additionally, extending the reaction scope to new sulfonamides while using preformed reagents is limited by the necessity to synthesize a different reagent for each case. In this work, we sought to explore the potentially broader amide substrate scope offered for direct α-C–H amidation of alkyl ethers using the simple sulfonamide/PhI(OAc)2 system.
![]() | (2) |
<!--Col Count:6-- > Entry | Oxidant | Solvent | T (°C) | t (h) | Yieldb (%) |
---|---|---|---|---|---|
a Conditions: 1a (0.2 mmol, 46 mg), oxidant (0.4 mmol, 2 equiv.), THF or mixed solvent (2 mL), reactions carried out under N2 atmosphere.b Yields determined by 1H NMR spectroscopy using trimethoxybenzene as internal standard.c 1 Equiv. of PhI(OAc)2.d 1.5 Equiv. of PhI(OAc)2.e 3 Equiv. of PhI(OAc)2.f Chloramine-T was used instead of the pair 1a/PhI(OAc)2.g The -NTs adduct was detected in this case. | |||||
1 | PhI(OAc)2 | Neat | 25 | 10 | 53 |
2 | PhI(OAc)2c | Neat | 60 | 2 | 77 |
3 | PhI(OAc)2d | Neat | 60 | 2 | 89 |
4 | PhI(OAc)2 | Neat | 60 | 2 | 91 |
5 | PhI(OAc)2e | Neat | 60 | 2 | 94 |
6 | PhI(OPiv)2 | Neat | 60 | 2 | 93 |
7 | PhI(TFA)2 | Neat | 60 | 2 | 23 |
8 | Chloramine-Tf | Neat | 60 | 2 | 3g |
9 | NH4IO3 | Neat | 60 | 2 | <1 |
10 | KIO3 | Neat | 60 | 2 | <1 |
11 | KIO4 | Neat | 60 | 2 | <1 |
12 | nBu4NIO4 | Neat | 60 | 2 | <1 |
13 | PhI(OAc)2 | DCE/THF (1![]() ![]() |
60 | 4 | 71 |
14 | PhI(OAc)2 | MeCN/THF (1![]() ![]() |
60 | 4 | 83 |
15 | PhI(OAc)2 | Toluene/THF (1![]() ![]() |
60 | 4 | 42 |
16 | PhI(OAc)2 | Cyclohexane/THF (1![]() ![]() |
60 | 4 | 56 |
In terms of the sulfonamide scope, we obtained excellent yields with both aromatic and aliphatic sulfonamides (Table 2). This constitutes an important benefit compared to the previous metal-free amidation of ethers reported by Ochiai, which, despite its excellent activity, was limited to the highly reactive p-(CF3)(Ph)BrNTf reagent.34 In our system, α-CH amidation of THF takes place with excellent yields for substituted and unsubstituted aromatic sulfonamides (entries 1–5), although no conversion was detected in the case of the –OMe and –NH2 derivatives (entries 6 and 7). Aliphatic sulfonamides also gave excellent conversions and high selectivity for the α-CH position (entries 8–10). The broad sulfonamide scope exhibited by our method is especially significant considering the useful biological and pharmacological properties of molecules based on the N-(methoxoalkyl)sulfonamide framework.51–54
<!--Col Count:4-- > Entry | Substrate | –R | Yieldb (%) |
---|---|---|---|
a Conditions: 1b–k (0.2 mmol), oxidant (0.3 mmol, 1.5 equiv.), neat THF (2 mL), 60 °C, 2 h, reactions carried out under N2 atmosphere.b Yields were determined by 1H NMR spectroscopy using trimethoxybenzene as internal standard. | |||
1 | 1b | –C6H5 | 89 |
2 | 1c | –C6H4(p-Me) | 90 |
3 | 1d | –C6H4(p-Cl) | 92 |
4 | 1e | –C6H4(p-Br) | 95 |
5 | 1f | –C6H4(p-COOH) | 71 |
6 | 1g | –C6H4(p-OMe) | <1 |
7 | 1h | –C6H4(p-NH2) | <1 |
8 | 1i | –(CH2)Ph | 92 |
9 | 1j | –CH3 | 89 |
10 | 1k | –CF3 | 96 |
The new functionalized THF compounds 2a–k were characterized by 1H and 13C{1H} NMR spectroscopy as well as high-resolution mass spectrometry (see ESI†). A characteristic 1H NMR resonance in the range from 5.20 to 5.42 ppm is assigned to the O–CH–NH proton and is common to all compounds 2a–k; the corresponding 13C{1H} signal appears at ca. 85 ppm. The molecular structures of two of these species (2a and 2d) were further confirmed by X-ray diffraction studies (Fig. 1). Their C–N bond distances are identical within the experimental error, with values of 1.472(8) Å and 1.470(4) Å for 2a and 2d, respectively, which are consistent with single bond character. Different rotameric orientations of the THF ring are present in each structure, likely due to packing forces.
![]() | ||
Fig. 1 ORTEPs of compounds 2a and 2d. 50% Thermal ellipsoids are shown. Most hydrogen atoms have been omitted for clarity. |
Under our optimized conditions (60 °C, 2 h, 1.5–2 equiv. PhI(OAc)2), a series of aliphatic and aromatic, cyclic and acyclic ethers were tested as substrates for amidation with 1a/PhI(OAc)2; however, the results were unsatisfactory in terms of yields (from 5 to 10%) and selectivity, except for the cyclic tetrahydropyran (THP, 50% yield). Both the less polar character of acyclic ethers compared to THF and their slightly higher α-C–H bond dissociation energies55 might contribute to the decreased activity. Further optimization of reaction conditions for diethyl ether (Table S1 in the ESI†) resulted in a maximum yield of 35% for the desired amidated product, but its formation was accompanied by a number of other unidentified species (60% overall sulfonamide conversion). Interestingly, we observed improved activity and considerably enhanced selectivity under microwave irradiation (MW, 120 °C, 10 min) in the neat ethers (Table 3, entries 4, 5; and Table S2†). We carried out the same reactions in pressurized microwave vials but heated in a conventional oil bath (Table 3, yields in parentheses).56–58 The results were almost identical, suggesting a thermal origin for the improved yields and selectivity under microwave irradiation. For the sake of comparison, we undertook the α-CH amidation of THF under MW conditions and obtained yields comparable to those observed under mild thermal heating.
<!--Col Count:6-- > Entry | Substrate | Product | Additive (equiv.) | Yieldb (%) | Selectivityc |
---|---|---|---|---|---|
a Conditions: 1a (0.5 mmol), PhIO(Ac)2 (0.55 mmol, 1.1 equiv.), neat ether (2 mL), microwave radiation (MW): 120 °C, 10 min. Reactions carried out under N2 atmosphere.b Yields were determined by 1H NMR spectroscopy using trimethoxybenzene as internal standard. Values between parentheses correspond to yields obtained under identical conditions as for MW experiments but heating in conventional oil bath.c Selectivities were calculated for sulfonamide as the limiting reagent.d These yields are calculated based on the ether (instead of the sulfonamide) as the limiting reagent, since the reaction also proceeds in the absence of 1a. | |||||
1 | ![]() |
![]() |
— | 62 (56) | 0.82 (0.79) |
2 | I2 (0.02) | 59 | 0.76 | ||
3 | I2 (0.5) | 3 | 0.20 | ||
4 | ![]() |
![]() |
— | 60 (51) | 0.83 (0.78) |
5 | I2 (0.02) | 70 | 0.89 | ||
6 | ![]() |
![]() |
— | 27 (27) | 0.49 (0.44) |
7 | I2 (0.02) | 68 | 0.84 | ||
8 | I2 (0.5) | <1 | 0.10 | ||
9 | ![]() |
![]() |
— | 12 (12) | 0.21 (0.24) |
10 | I2 (0.02) | 54 | 0.86 | ||
11 | ![]() |
![]() |
— | 14 | — |
12 | I2 (0.02) | 26 | — | ||
13 | ![]() |
![]() |
— | 79 (85) | 0.98 (0.94) |
14 | I2 (0.02) | 60 | 0.83 | ||
15 | I2 (0.5) | 15 | 0.30 | ||
16 | ![]() |
![]() |
— | 14 (13) | 0.61 (0.54) |
17 | I2 (0.02) | 3 | 0.10 | ||
18 | ![]() |
![]() |
— | 10, 1.5d, 11, 1.9d | — |
19 | ![]() |
![]() |
— | 10, 5.6d, 11, 3.5d | — |
The addition of iodine in catalytic or stoichiometric amounts in conjunction with hypervalent iodine oxidants has been widely and successfully employed in other C–H amidation reactions.24,29,35–37 We observed an interesting effect with the addition of variable amounts of iodine. In the presence of catalytic amounts (2 mol% I2 relative to sulfonamide) the yields for linear acyclic ethers increased considerably. For instance, the yield for di-n-butyl ether increased from 12 to 54% (entries 9 and 10), whereas that of di-n-propyl ether increased from 27 to 68% (entries 6 and 7). However, amidation of cyclic ethers (THF and THP) was not affected by the presence of trace amounts of iodine, which even had a deleterious effect in the case of tert-butyl ethers. In all cases, higher amounts of iodine (0.5 equiv.) greatly suppressed product formation. In contrast, 0.5 equiv. of I2 was previously found to be the optimal loading for benzylic amidations by Fan and coworkers who obtained lower yields at both 1 equiv. of I2 and 0.05 to 0.2 equiv. of I2.24 These results highlight the importance of testing a wide range of I2 concentrations in optimizing this type of reaction.
While linear cyclic and acyclic ethers gave the expected α-amidated products, both aromatic and branched aliphatic ethers exhibited a different reactivity. The branched di-isopropyl ether could not be amidated under our experimental conditions; instead we observed the formation of isopropyl acetate (7) as the major product, due to direct acetoxylation, determined by 1H NMR after addition of an authentic sample of 7. In contrast, acetoxylation was barely detected with the other ether substrates. Surprisingly, reaction of tert-butyl methyl ether and tert-butyl ethyl ether gave completely different outcomes. The former yielded compound 8 where the O-methyl fragment is doubly functionalized and the tert-butoxy groups released. The molecular structure of 8 is consistent with the reduced number of resonances appearing in its 1H and 13C{1H} NMR spectra and was confirmed by X-ray diffraction (Fig. 2). On the other hand, tert-butyl ethyl ether was converted into compound 9 after α-CH amidation of the ethyl group and unexpected β-CH amidation of the tert-butyl fragment. This species was unambiguously characterized by 1H and 13C{1H} NMR spectroscopy, as well as HRMS (FT-ICR). It constitutes a rare example of amidation of unactivated primary C–H bonds. A somewhat related reaction has been recently reported by Che and coworkers, where the C–H bond of a tert-butyl fragment was also aminated, although in low yield as a result of competition with the more reactive benzylic C–H bond.59 Finally, and in contrast to the related copper-catalyzed process,60 we observed that benzylic ethers cannot be amidated using the present method; instead they experience C–O cleavage and further oxidation to yield benzaldehyde and benzoic acid. The same transformation has previously been reported under related conditions and mainly investigated as a protocol for the synthesis of aldehydes from benzylic ethers.61–63 Although these reactions were accelerated by the addition of the sulfonamide and especially by the presence of the hypervalent iodine oxidant, they proceed even in the absence of these additives (see ESI for more details, Table S3†).
![]() | ||
Fig. 2 ORTEPs of compounds 8 and 12. 50% Thermal ellipsoids are shown. Most hydrogen atoms have been omitted for clarity. |
Purification of the functionalized ethers reported in this work proved difficult. In most cases extensive decomposition occurred during attempted column chromatography on both silica gel and alumina. In some cases degradation of the amidated compounds was observed after mild work-up or even upon standing in air for several days. The intrinsic reactivity of these species gave rise to an interesting derivative (12, Scheme 2), which is formed from Friedel–Crafts alkylation of electron-rich trimethoxybenzene (employed as an internal standard for1 H NMR spectroscopy analysis). The Friedel–Crafts coupling of sulfonamides with aromatic compounds typically employs N-sulfonyl aldimines,64–67 but the direct use of N,O-acetals has been less explored. In an elegant study, Du Bois and co-workers developed a related rhodium-catalyzed one-pot amidation/nucleophilic addition process using ethers, sulfonamides and nucleophiles such as allylsilanes, silyl enol ethers, and ketene acetals.68 Another interesting related process for the synthesis of 3-oxyindoles via tandem amidation/acetoxylation/Friedel–Crafts reaction has been recently described.69 The molecular structure of 12 was elucidated by X-ray diffraction studies of suitable crystals grown by slow diffusion of pentane into a dichloromethane solution of the compound (Fig. 2). The length of the new C–C bond, 1.512(3) Å, and that of the adjacent C–N bond, 1.490(3) Å, are both indicative of single bonds.
Attempts to perform amination reactions of primary and secondary alkyl amines with THF using PhI(OAc)2 led to acetoxylation product (compound 14 in Scheme 3) as evidenced by1 H and 13C{1H} NMR spectroscopy in comparison with literature data.70 This reaction is presumably base-mediated as the yields were dependent on the amount of amine present, and the same product could be formed using inorganic base K2CO3 in place of the amine.
To investigate the feasibility of the classic mechanism for in situ nitrene formation (Scheme 1), which requires a primary sulfonamide,6,10,11 we tested the reactivity of methyl tosylamide with THF. Interestingly, we found that amidated product is formed in 14% yield along with 17% of the acetoxylation product (Scheme 3). Compound 13 displays 1H and 13C{1H} NMR spectra similar to a previously reported and related product derived from cyclohexyl tosylamide.71 This result suggests that in situ nitrene formation is not the only available pathway in the present system, and there is at least one other mechanism implicated that would account for the amidation using secondary sulfonamides.
The amidation of THF by sulfonamide 1a was completely shut down by addition of the radical trap TEMPO, suggesting a mechanism involving radicals. Two distinct radical-based mechanisms for PhI(OAc)2-based systems have been proposed in the literature for oxidative C–N functionalization of sp3 benzylic C–H bonds. In an account by Cho and Chang involving imidation of benzylic C–H bonds using PhI(OAc)2 alone, a carbon-centered radical R1(˙CH)R2 at the benzylic position was proposed from a single-electron transfer (SET) reaction between the benzylic substrate R1(CH2)R2 (R1 = Ar) and PhI(OAc)2 (Scheme 4, pathway A).23 A second SET to form a carbocation R1(+CH)R2; followed by attack by a sulfonimide nitrogen and removal of a proton would then produce the final amidation product. Accounts describing amidation and amination of sp3 carbons using PhI(OAc)2/I2 systems are in general agreement in proposing mechanisms involving formation of a nitrogen-centered R3R4N˙ radical resulting from reaction of the oxidant/I2 combination with substrate either thermally or by photoactivation.24,29,36,37 Subsequent hydrogen atom transfer (HAT) produces a carbon-centered radical R1(˙CH)R2, which abstracts iodine from R3R4NI to propagate the production of the R3R4N˙ radical. Substitution of the R1(CH)R2 I species with either R3R4NH or AcOH can lead to formation of amination/amidation product or the undesired acetoxylation byproduct.
In our system, a substantial KIE of kH/D = 9.3 is observed for the reaction of 1a in a 1:
1 mixture of THF and THF-d8 with 2 equiv. of PhI(OAc)2. Under identical conditions after addition of 2 mol% I2, the kH/D is decreased to 2.2. The large difference in KIE between these two cases implies a change in mechanism which may involve a shift from pathway A operative with PhI(OAc)2 alone to pathway B upon addition of I2 (Scheme 4). Future studies will focus on further elucidating the reaction mechanism(s) implicated and understanding the dependence of the role of I2 on the identity of the ether substrate.
Footnotes |
† Electronic supplementary information (ESI) available: Experimental procedures, 1H and 13C NMR data for all new compounds, additional optimization data, and crystallographic details for compounds 2a, 2d, 8 and 12. See DOI: 10.1039/c4ra09665g |
‡ These authors contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2014 |