Visible-Light Mediated Oxidative Fragmentation of Ethers and Acetals by Means of Fe(III) Catalysis

A new method employing iron(III) acetylacetonate along with visible light is described to effect oxidative ring opening of cyclic ethers and acetals with unparalleled efficiency. The method allows for a photocatalytic radical chemistry approach to functionalize relatively inert cyclic ethers into useful synthetic intermediates. The methodology sheds further light on the use of underexplored iron complexes in visible-light photochemical contexts and illustrates that simple Fe(III) complexes can initiate redox processes from 4LMCT excited states.


Deviation from standard conditions
The reactions were prepared according to the procedure described in the section Experimental procedure for photoreaction. S-5

Screening of metal additives
The reactions were prepared according to the procedure described in the section Experimental procedure for photoreaction. The reactions were analyzed by taking 0.1 ml of crude reaction mixture and diluted with 0.6 ml CDCl3 followed by recording 1 H-NMR, no prior work-up. Results are summarized below.
No. Metal additive Conversion (%) Fe(acac)3 100 Table 2. 1 mol% of the metal additive and 2-(4-chlorophenyl)tetrahydrofuran (1aa) were used in each reaction and run for 6 hours. All vials and stir bars were carefully cleaned with aqua regia before reaction.
To double-check that Fe(acac)3 and Co(ClO4)2 are active catalysts for the reaction experiments were conducted again for these two metal complexes in brand new vials. In case of Fe(acac)3 the results were always reproducible while for Co(ClO4)2 slightly diminished yields were sometimes obtained. Hence, we concluded that Fe(acac)3 was the most efficient catalyst and continued exploring it in our system.

Synthesis of starting materials
Two general methods have been used to prepare a large portion of the starting materials. The specific syntheses of starting materials beyond these two methods are described in the following section together with the characterization data for that compound.

Method I -THF-derivatives
The corresponding 4-oxo-4-phenylbutyric acid (5 mmol) was added portion wise to a suspension of LiAlH4 (4 eq.) in Et2O (50 ml) at 0 o C. Reaction was then stirred at r.t until completed as judged by TLC. Work-up procedure according to the Fieser protocol. Almost quantitative yields of the diols were obtained, and the diols were used in the next step without further purification. The cyclization of the diols to tetrahydrofurans was conducted as described in [1] . No purification on silica was necessary.

Method II -THF, THP and dioxane-derivatives
Cyclic ethers were synthesized according to procedure in [2] . Oven dried Biotage MW vial (20 ml) was charged with the boronic acid (2 mmol), Ni(acac)2 (10 mol%, 51 mg), PPh3 (10 mol%, 52 mg), K3PO4 (1 eq., 425 mg) and THF (anhydrous, 12 ml). The vial was capped and reaction mixture sparged with argon for 10 min. Then DTBP (1.2 eq., 0.44 ml) was added in one portion and the reaction mixture heated to 100 o C overnight. After cooled to r.t the reaction was diluted with 20 ml HCl(aq) (2M) 1 and extracted 3 times with 20 ml Et2O, dried over Na2SO4 and concentrated under reduced pressure. Products were then purified on silica. Seebach's stain turned out to be superior in visualizing the product on TLC plates.
Care must be taken regarding reduced pressure since the compounds are, in some cases, volatile. 20 mbar and 40 o C water bath worked fine in most cases, exception being the 2,4-difluoro (compound X).  [5] .  [4] .

S-12
(3jj(a)) Benzylmethyl ether: To a solution of benzaldehyde (2 mmol, 212 mg) in 20 ml MeOH was added NaBH4 (1.2 eq. 91 mg). Reaction was stirred until completed as judged by TLC. Reaction mixture was concentrated under reduced pressure and redissolved in Et2O and washed with H2O, dried over Na2SO4 and concentrated under reduced pressure. Without further purification, the benzyl alcohol was dissolved in 20 ml anhydrous THF and NaH (1.5 eq. 60% dispersion in paraffin liquid, 120 mg) was added portionwise and stirred for 10 minutes. Then methyl iodide (5 eq., 0.62 ml) was added in one portion and stirred overnight. The reaction mixture was adsorbed onto Celite and purified on silica Rf(5% EtOAc in hexane) = 0.28, colorless oil (110 mg, 45% yield over two steps).

Experimental procedure for photoreaction
All reactions were run in duplicate unless otherwise stated. Dichloroethane (DCE, anhydrous) and BrCCl3 were deoxygenated by sparging with argon for 20 min. Cyclic ether or acetal (0.2 mmol), Fe(acac)3 (1 mol%) and a stir bar was added to a Biotage MW vial (10 ml) and capped. Reaction atmosphere was exchanged to argon by "sparging" head space for 10 min. Then BrCCl3 (3 eq. 60 µL) followed by DCE (0.1 M, 2 ml) was added. The capped end of the vial was wrapped with parafilm and then placed in a photoreactor and irradiated with 455 nm light for 18 h at 27 o C (with fan). The two duplicated reactions were combined and adsorbed onto Celite and purified on silica. Isolated yields were thereby obtained as an average.

UV-Vis
To test whether Fe(acac)3 and BrCCl3 could form an electron-donor-acceptor complex (EDA complex) UV-Vis were recorded for the two pure substances in DCE as well as their mixtures. A solution of Fe(acac)3 and BrCCl3 in DCE (0.2 mM and 60 mM respectively) was initially measured. BrCCl3 (300 eq. 17.7 µL) was then added to the cuvette and UV-Vis was once more measured (Mixture 1). A second equal portion of BrCCl3 was added to the same cuvette and UV-Vis recorded once more (Mixture 2). In the plot below the concentration has been corrected for the added volume of BrCCl3. As can be seen in Figure S1 below, no shift of the bands has occurred upon mixing Fe(acac)3 and BrCCl3 and no new charge-transfer bands can be seen. The only difference is a slight decrease in absorbance for Mixture 1 and Mixture 2, which correlates with the slightly diluted solution by the added volume of BrCCl3.

Reactions
Reactions in this section refers to those in Scheme 5 and Scheme 6. All reaction were prepared as described in Experimental procedure for photoreaction and conducted at 0.1 mmol scale with respect to 1aa. Experiment i): 1aa, 1 eq. Fe(acac)3 and no BrCCl3 in DCE, irradiated for 6 h. These conditions did not provide any product.

Calculations on BrCCl3
Geometry optimization was performed with the method CCSD+PCM/cc-pVTZ, solvent: dichloroethane.